Information

In insects, does the Alanine repeat occur on the homeodomain sequence of the abdomen or does it occur on a different sequence?


By "Alanine repeat", I am referring to the suppression of the formation of extra insect legs due to Ubx gene suppression through Distal-less repression.


Some background:

The new work involved misexpression of the Drosophila Ubx protein in the presumptive thorax of transgenic fruitfly embryos. Limb development was suppressed because of repression of Dll. By contrast, the misexpression of onychophoran and crustacean Ubx proteins did not interfere with Dll expression and the formation of thoracic limbs. These results raised the possibility that the Drosophila Ubx protein is functionally distinct from Ubx in onychophorans and crustaceans. One study suggests that Drosophila Ubx has acquired an alanine-rich peptide that mediates the repression of gene transcription; this peptide is lacking in onychophorans. The other study provides evidence that the crustacean Ubx contains an additional peptide that modulates the activity of the alanine-rich peptide, and possibly other repression domains, in crustacean Ubx.

Insect-restricted sequences include four regions N-terminal to the homeodomain (I1-I4), a peptide motif (QAQAQK), and an extended run of alanine residues C-terminal to the homeodomain

The onychophoran Ubx protein might function as an activator of appendage development. When the onychophorans and arthropods diverged, Ubx acquired an alanine-rich repression domain near its carboxy terminus. This domain mediates constitutive repression in insects. But in crustaceans the addition of the regulatory peptide causes it to function in a conditional fashion. As a result, Ubx does not suppress limb development in crustaceans. But it eliminates abdominal limbs in insects, greatly reducing the overall number of appendages compared with crustaceans.

You can read about it here:

Levine, Mike. "Evolutionary biology: how insects lose their limbs." Nature 415.6874 (2002): 848-849.

Galant, Ron, and Sean B. Carroll. "Evolution of a transcriptional repression domain in an insect Hox protein." Nature 415.6874 (2002): 910-913.


Tandem amino acid repeats in the green anole (Anolis carolinensis) and other squamates may have a role in increasing genetic variability

Tandem amino acid repeats are characterised by the consecutive recurrence of a single amino acid. They exhibit high rates of length mutations in addition to point mutations and have been proposed to be involved in genetic plasticity. Squamate reptiles (lizards and snakes) diversify in both morphology and physiology. The underlying mechanism is yet to be understood. In a previous phylogenomic analysis of reptiles, the density of tandem repeats in an anole lizard diverged heavily from that of the other reptiles. To gain further insight into the tandem amino acid repeats in squamates, we analysed the repeat content in the green anole (Anolis carolinensis) proteome and compared the amino acid repeats in a large orthologous protein data set from six vertebrates (the Western clawed frog, the green anole, the Chinese softshell turtle, the zebra finch, mouse and human).

Results

Our results revealed that the number of amino acid repeats in the green anole exceeded those found in the other five species studied. Species-only repeats were found in high proportion in the green anole but not in the other five species, suggesting that the green anole had gained many amino acid repeats in either the Anolis or the squamate lineage. Since the amino acid repeat containing genes in the green anole were highly enriched in genes related to transcription and development, an important family of developmental genes, i.e., the Hox family, was further studied in a wide collection of squamates. Abundant amino acid repeats were also observed, implying the general high tolerance of amino acid repeats in squamates. A particular enrichment of amino acid repeats was observed in the central class Hox genes that are known to be responsible for defining cervical to lumbar regions.

Conclusions

Our study suggests that the abundant amino acid repeats in the green anole, and possibly in other squamates, may play a role in increasing the genetic variability, and contribute to the evolutionary diversity of this clade.


Recognition of non-self

Immune reactions are initiated when molecules of microbial origin are detected and recognized as `non-self'. This recognition step involves pattern recognition receptors (PRRs) that recognize and bind to so-called pathogen-associated molecular patterns (PAMPs) that are shared by various microorganisms but absent from eukaryotic cells (Medzhitov and Janeway, 1997b, 2002). PAMPs that are known to be potent immune elicitors include lipopolysaccharides (LPS), peptidoglycan(PGN) and β-1,3-glucans. Various PRRs have been identified and isolated from vertebrates (Holmskov et al.,2003 Takeda et al.,2003) and invertebrates(Gobert et al., 2003 Hoffmann, 2003 Wilson et al., 1999 Yu et al., 2002). The best-studied invertebrate PRRs are the peptidoglycan recognition proteins(PGRPs) and the Gram-negative bacteria-binding proteins (GNBPs). PGRPs are soluble or transmembrane proteins containing a domain similar to the bacterial amidase domain, which is involved in recycling bacterial cell wall fragments. PGRPs have been isolated from both invertebrates and vertebrates(Dziarski et al., 2003 Wang et al., 2003). The first was isolated from the haemolymph of the moth Bombyx mori, where it is involved in activating the prophenoloxidase (PPO) cascade(Yoshida et al., 1996). In Drosophila, three PGRPs were identified as bona fide PRRs. The soluble PGRP-SA activates the Toll pathway in response to Gram-positive bacterial infection (Gobert et al.,2003 Michel et al.,2001) in concert with another PRR, GNBP1. By contrast, PGRP-LC(Choe et al., 2002 Gottar et al., 2002 Ramet et al., 2002) and PGRP-LE (Takehana et al.,2002) are involved in activating the immune deficiency (IMD)pathway in response to Gram-negative bacterial infections. Among the seven identified putative Anopheles PGRPs(Christophides et al., 2002),PGRPLC seems to play a central role in defence against bacterial infection (G. K. Christophides, unpublished data). The orthologous genes in both Drosophila (Werner et al.,2003) and Anopheles(Christophides et al., 2002)undergo alternative splicing resulting in at least three distinct isoforms. Interestingly, the Anopheles PGRPLC isoforms are differentially regulated upon immune challenge, suggesting that splicing can be regulated by the immune signal (Christophides et al.,2002). On the other hand, PGRPLB is transcriptionally upregulated following Plasmodium infection of adult mosquitoes(Dimopoulos et al., 2002) and maintains an elevated expression throughout the parasite's entire life cycle(Christophides et al.,2002).

GNBPs were first described in Bombyx mori(Lee et al., 1996) and share significant sequence similarity with the catalytic region of bacterialβ-1,3- and β-1,3,1,4-glucanases. BmGNBP binds strongly to the surface of Gram-negative bacteria and shows pronounced transcriptional upregulation following bacterial challenge(Lee et al., 1996). The Drosophila GNBP1 binds with high affinity to LPS andβ-1,3-glucan (Kim et al.,2000) and, in concert with PGRP-SA, activates the Toll pathway upon infection with Gram-positive bacteria(Gobert et al., 2003). In A. gambiae, six putative GNBPs have been identified(Christophides et al., 2002). Among them, GNBPB1 and GNBPA1 are upregulated following Plasmodium infection, while only GNBPB1 is responsive to bacteria (Christophides et al.,2002 Dimopoulos et al.,2002). Other putative Anopheles PRRs include the thioester-containing proteins (TEPs), leucine-rich immune proteins (LRIMs) and C-type lectins (CTLs). Members of these protein families were recently implicated in the regulation of Plasmodium development in the mosquito vector and are discussed later in this review.


Results

Conservation of Orthologous TpnI and TpnT Gene Structures in Insects

We have used the previously reported D. melanogaster TpnT and TpnI gene sequences ( Barbas et al. 1993 Benoist et al. 1998) to find putative orthologues in the Anopheles and Apis genomes. The structural organization of these genes has been preserved to a high degree in insects ( fig. 1). In general, we observe an increase in gene size, mainly due to intron length expansion from the smallest genome—178 Mb in D. melanogaster ( Adams et al. 2000)—to those of Anopheles and Apis that are almost double in size ( Holt et al. 2002). Chromosomal location is not a broadly preserved feature in insects, neither for TpnC genes ( Herranz, Mateos, and Marco 2005) nor for those of TpnI or TpnT. The TpnT and TpnI genes are both located on the D. melanogaster X chromosome, but the Anopheles TpnT gene is located cytologically on chromosome arm 2R, while TpnI remains to be localized ( Mongin et al. 2004).

Troponin T (top) and troponin I (bottom) gene structures in holometabolous insects. Genes are labeled with the species binomial two-letter abbreviation. Noncoding (white), constitutive (gray), alternative (black) or mutually exclusive (diagonal stripes), and the PAANGKA- or APPAEGA-containing (horizontal stripes) alternative exons are indicated. Light shaded exon 6's in TpnI have not yet been detected in RT-PCR experiments. Exon numbers in Anopheles and Apis genes have been assigned by homology to those of their Drosophila counterparts. Genomic DNA is represented by a continuous line that is therefore missing in gene sequences derived from cDNA data. Gene organization is well preserved overall, and gene size correlates with the organism genome size. The main difference affects alanine/proline-rich sequences including the TpnI exon 3 encoding the PAANGKA sequences in the Drosophilidae, but not in Anopheles or Apis. Several Apis TpnI gene exons encode an APPAEGA-containing sequence in their 3′ region.

Troponin T (top) and troponin I (bottom) gene structures in holometabolous insects. Genes are labeled with the species binomial two-letter abbreviation. Noncoding (white), constitutive (gray), alternative (black) or mutually exclusive (diagonal stripes), and the PAANGKA- or APPAEGA-containing (horizontal stripes) alternative exons are indicated. Light shaded exon 6's in TpnI have not yet been detected in RT-PCR experiments. Exon numbers in Anopheles and Apis genes have been assigned by homology to those of their Drosophila counterparts. Genomic DNA is represented by a continuous line that is therefore missing in gene sequences derived from cDNA data. Gene organization is well preserved overall, and gene size correlates with the organism genome size. The main difference affects alanine/proline-rich sequences including the TpnI exon 3 encoding the PAANGKA sequences in the Drosophilidae, but not in Anopheles or Apis. Several Apis TpnI gene exons encode an APPAEGA-containing sequence in their 3′ region.

We have cloned and sequenced the transcripts of these genes in the drosophilid species and Apis, paying special attention to the alternatively spliced exons. In the case of troponin T, the differentially spliced N t exons are present in all the analyzed species, although in Apis the TpnT gene shows an additional alternatively spliced exon, named 5′. The Apis equivalent to the dipteran exon 6 is divided into five constitutive exons (a, b, c, d, and e). Sequencing of transcripts identified a new variable exon 10, that had been overlooked in previous studies of D. melanogaster and D. virilis sequenced TpnT transcripts ( Benoist et al. 1998). Comparison of the TpnT genomic sequences in insects showed that two different exons, named 10A and 10B, are present in all holometabolous species. Both are 79-nt long, encoding 26 aa. Another difference in the TpnT gene structure is found in exon 11, which encodes a long polyglutamic tail with variable size in all studied protostome species ( Benoist et al. 1998).

The constitutive exons and exon 9 of TpnI gene are present in all analyzed insects ( fig. 1), but exon 3 is not, to the extent that the Drosophila orthologous exons 2 and 4 are fused in a unique exon in Anopheles, but not in Apis where a smaller exon 3 is found the main variability of Troponin I arises from mutually exclusive exon 6's. In the case of Anopheles, we detected three exon 6 variants by sequence homology, while four were detected in Apis. Interestingly, additional exons, that we have designated H1, H2, and H3, were found downstream of exon 10 in the Apis TpnI gene, encoding the APPAEGA-repetitive sequences (see below).

Sequence Variations of the Alternatively Spliced Exons in TpnT and TpnI Genes

Amino acid sequences from TpnT and TpnI constitutive exons are conserved almost perfectly in insects. As shown in figure 2A, even the sequences of the alternatively spliced exons (3, 4, and 5) from TpnT are well conserved among the Drosophilidae. Exons 3 and 4 are also conserved in A. gambiae and A. mellifera while a higher degree of variation is found in exon 5, reflecting the evolutionary distance between these insects. In this region, the Apis TpnT gene has an additional alternatively spliced exon. Exons 3 and 4 are clearly conserved, but exons 5, 5′, and the constitutive exon 6a have very different sequences, being more similar to sequences in the orthopteran Periplaneta ( Wolf 1999) and the odonate Libellula TpnT exons ( Fitzhugh and Marden 1997 Marden et al. 1999, 2001) despite the much larger evolutionary distance of these species from holometabolous insects.

Alternatively spliced exons in TpnT and TpnI genes. (A) Alignment of the 5′ region of the troponin T transcripts in insects. Amino acid conservation (•), silent variation () (cDNA), equivalent variation (), and nonequivalent variation () in relation to the Drosophila melanogaster sequence are shown. These sequences, labeled with the two-letter species abbreviations and the corresponding exon number (E10A Dm means exon 10A in D. melanogaster for instance), are conserved in the four drosophilid species (only silent variation appears) and also in the Anopheles and Apis exons except exon 5. The Libellula (AF133521) and Periplaneta (AF133520) TpnT sequences (obtained from GenBank) retain the same properties although the alternative exon sequences show higher levels of variation. (B) Alignment and phylogenetic tree of the alternative TpnT exon 10 sequences in insects. Branches with bootstrap values below 70% have been collapsed in this cDNA-based tree. Most changes affect the number of phosphorylable residues, indicated at the right of the alignment, with the exception of the conserved final threonine marked with an asterisk. (C) Alignment and phylogenetic tree of the alternative TpnI exon 9 and 10 sequences in insects. The Haemaphysalis longicornis TpnI sequence (AB051079), in which exon 9 has not been described, was used as out-group.

Alternatively spliced exons in TpnT and TpnI genes. (A) Alignment of the 5′ region of the troponin T transcripts in insects. Amino acid conservation (•), silent variation () (cDNA), equivalent variation (), and nonequivalent variation () in relation to the Drosophila melanogaster sequence are shown. These sequences, labeled with the two-letter species abbreviations and the corresponding exon number (E10A Dm means exon 10A in D. melanogaster for instance), are conserved in the four drosophilid species (only silent variation appears) and also in the Anopheles and Apis exons except exon 5. The Libellula (AF133521) and Periplaneta (AF133520) TpnT sequences (obtained from GenBank) retain the same properties although the alternative exon sequences show higher levels of variation. (B) Alignment and phylogenetic tree of the alternative TpnT exon 10 sequences in insects. Branches with bootstrap values below 70% have been collapsed in this cDNA-based tree. Most changes affect the number of phosphorylable residues, indicated at the right of the alignment, with the exception of the conserved final threonine marked with an asterisk. (C) Alignment and phylogenetic tree of the alternative TpnI exon 9 and 10 sequences in insects. The Haemaphysalis longicornis TpnI sequence (AB051079), in which exon 9 has not been described, was used as out-group.

The sequences of the mutually exclusive TpnT exons 10A and 10B ( fig. 2B) are also well preserved among Drosophilidae. Larger differences are found in Anopheles and Apis. Nonequivalent variations in exon 10A are equally distributed throughout the exon, but less so in exon 10B. Interestingly, a higher number of threonine residues appear in insect exon 10A than in exon 10B, but only the last threonine of exon 10A is conserved in all the insects analyzed. Drosophila TpnT isoforms are readily phosphorylated in vivo ( Domingo et al. 1998), precisely in the exon 10A–encoded sequence (Nongthomba and Sparrow, personal communication).

With regard to the TpnI genes, constitutive exons are almost strictly conserved and the sequence of alternative exons 9 and 10 ( fig. 2C) reflect a situation similar to that of TpnT exons 10A and 10B. Even though we know that the alternatively spliced sequences are short, phylogenetic trees based on the comparison among the sequences encoded by these exons can be constructed. Although this information is not sufficient to establish the evolutionary origin of these exons, it can be used to detect how they have been changing in different insect groups when the data of all exons are analyzed together. In both Tpn genes, one of the duplicated exons (TpnT exon 10A and TpnI exon 9) appears to be less conserved than the other (TpnT exon 10B and TpnI exon 10), showing similar evolutionary patterns in each of the three insect groups analyzed. It is therefore likely that the duplication of these TpnT and TpnI alternative exons may have occurred at the origin of the holometabolous-type of development. In accordance with this idea, only a single TpnT exon 10 with intermediate sequence features has been found in the hemimetabolous Periplaneta americana and Libellula pulchella. Furthermore, only a single exon 9/10 has been found in the arachnid Haemaphysalis longicornis TpnI gene ( You et al. 2001). Interestingly, TpnI exon 9 in the Drosophilidae does not contain a complete stop codon, which is formed by the splicing of its last nucleotide with the two first bases of exon 10 ( Barbas et al. 1993). This feature is not observed in Anopheles or Apis, where stop codons and 3′ UTRs are found both in exons 9 and 10.

The TpnI mutually exclusive exon 6 sequences have also been analyzed. In a consensus phylogenetic tree based on exon 6 nucleotide sequences ( fig. 3A), the low bootstrap values at the base of the insect tree suggest an ancient origin for these duplication events, but the tree supports a clear differentiation of exon 6a's from exon 6b's. The 6b exons are variable enough to be phylogenetically discriminative. The 6a exons are so different among themselves that the Apis paralogous exons 6a1 and 6a2 are as related to each other as to their putative drosophilid orthologous exons (see the polytomy in the 6a exon branch). The presence of a single exon 6a in Anopheles and the loss of the exon 6a2 canonical donor-splicing signal in the Drosophilidae are also noteworthy. In a sequence alignment ( fig. 3B), it can be seen that each Anopheles and Apis exon 6 is as closely related to each other as it is to those from the Drosophilidae. Finally, the Drosophilidae TpnI exon 3 ( fig. 4A) are essentially alanine/proline-rich sequences. All show the PAANGKA motif which appears repeated several times at different positions in the D. melanogaster, D. subobscura, and D. virilis extensions. This domain is absent in Anopheles and in Apis TpnIs, where a very short constitutive exon 3 is found.

Phylogenetic tree and alignment of the differentially spliced TpnI exon 6's. (A) Branches with bootstrap values below 70% have been collapsed in this cDNA-based tree. Exons 6b1 and 6b2 are clearly separated, but an exon 6a phylogeny cannot be extracted from the tree. (B) Alignment of the translated exon 6 sequences (34 aa translated from the last nucleotide of exon 5 plus the exon 6 nucleotides), flanked by the splicing sequences when available. The sequences are designated by the two-letter species abbreviation and the corresponding exon number (Dm E6b1 for instance). The 6a exons are the more variable ones, having even lost the canonical splicing donor sequence (bold and italics letters in the figure) in the drosophilid 6a2 exons.

Phylogenetic tree and alignment of the differentially spliced TpnI exon 6's. (A) Branches with bootstrap values below 70% have been collapsed in this cDNA-based tree. Exons 6b1 and 6b2 are clearly separated, but an exon 6a phylogeny cannot be extracted from the tree. (B) Alignment of the translated exon 6 sequences (34 aa translated from the last nucleotide of exon 5 plus the exon 6 nucleotides), flanked by the splicing sequences when available. The sequences are designated by the two-letter species abbreviation and the corresponding exon number (Dm E6b1 for instance). The 6a exons are the more variable ones, having even lost the canonical splicing donor sequence (bold and italics letters in the figure) in the drosophilid 6a2 exons.

Alignment of the IFM-specific troponin I and tropomyosin alanine/proline-rich extensions. (A) Although the TpnI exon 3 sequences are not perfectly conserved even in the Drosophilidae, the heptad PAANGKA (shaded box) is conserved. It is repeated in some species surrounded by an alanine/proline-rich region. This sequence is absent in Anopheles and Apis because their TpnI genes lack an exon 3 with these properties. (B) The TpnH extension is encoded by a single Tm1 gene exon in Anopheles, while in the Apis TpnI gene it involves up to three gene 3′ exons (separated in the three rows of the alignment), all of them encoding an alanine/proline-rich sequence with the APPAEGA motif (shaded box). An equivalent sequence has been found in the Lethocerus TpnI gene (AJ621044) and is included as the final sequence in the alignment.

Alignment of the IFM-specific troponin I and tropomyosin alanine/proline-rich extensions. (A) Although the TpnI exon 3 sequences are not perfectly conserved even in the Drosophilidae, the heptad PAANGKA (shaded box) is conserved. It is repeated in some species surrounded by an alanine/proline-rich region. This sequence is absent in Anopheles and Apis because their TpnI genes lack an exon 3 with these properties. (B) The TpnH extension is encoded by a single Tm1 gene exon in Anopheles, while in the Apis TpnI gene it involves up to three gene 3′ exons (separated in the three rows of the alignment), all of them encoding an alanine/proline-rich sequence with the APPAEGA motif (shaded box). An equivalent sequence has been found in the Lethocerus TpnI gene (AJ621044) and is included as the final sequence in the alignment.

Three new exons have been located downstream from exon 10 in the Apis TpnI gene. These exons can be spliced together to produce an Apis-specific TpnI isoform containing a C t extension encoding a repetitive APPAEGA sequence, through a new GT-splicing donor site in exon 10, located 25 bp before the stop codon. This extension is clearly similar to the alanine/proline-rich extension of the large–molecular weight IFM-specific tropomyosin in the Diptera, both in sequence and protein length ( fig. 4B). Interestingly, the Apis tropomyosin Tm1 gene lacks this type of sequence (Mateos et al., unpublished results). A similar repetitive alanine/proline extension is found at the 3′ end of a TpnI gene transcript from the hemipteran Lethocerus ( Qiu et al. 2003 and sequence AJ621044).

Expression Pattern of TpnI and TpnT Genes Are Conserved in Drosophilidae

The expression profiles of TpnI and TpnT isoforms in D. subobscura and D. virilis were detected using RT-PCR techniques (unpublished data). All the transcripts previously identified in D. melanogaster ( Barbas et al. 1993 Benoist et al. 1998) were detected in the second instar larvae and late pupae stages of these drosophilids. Combinations of the troponin T alternative exons 3, 4, and 5 appeared in larvae or in adult abdominal muscle transcripts in patterns characteristic for each muscle. The major transcripts of the adult thoracic muscles do not contain any of these alternative exons. Troponin I exon 9 inclusion is found exclusively in transcripts of the adult musculature, being accompanied by exon 3 in the major thoracic transcript. The four exon 6 types were detected at varying levels in all the tissues studied, those containing exon 6b1 being the more highly expressed in IFM ( Barbas et al. 1993).

The discovery of the new TpnT exon 10A led to a study of its expression pattern. A single probe for constitutive exon 6 and four specific probes for exons 10A or 10B ( fig. 5A) were used for RT-PCR. In embryos, only the 10B exon-containing transcript was detected, while in whole adult RNAs, transcripts containing either 10A or 10B appear.

Differences in the expression profile of TpnT and TpnI transcripts in the Drosophila melanogaster musculature. Agarose 1.2% gel separations of RT-PCRs. Each transcript detected by PCR as a band is marked with an arrow labeled with the name of the amplified gene and its exon composition. (A) Expression pattern of mutually exclusive exon 10's detected in adult or embryonic RNA extractions with four exon 10A- or 10B-specific primers (10A1, 10A2, 10B1, and 10B2). (B) Patterns of expression in adult body parts (head, IFM, TDT, and abdomen). Using probes for TpnT exon 6 and 10A (10A2 probe) or 10B (10B1 probe), the exon 10 expression pattern is shown. Exon 10B is expressed in all muscles in the adult except in the IFM. Exon 10A is the only one expressed in the IFM, but it is also expressed in the TDT muscles. (C) Probes for the 3′ region of the TpnI gene were used in RT-PCRs (exons 8–9 and 8–10). Exon 9 uses the first two residues of exon 10 to generate a stop codon, so the exon 10 probe detects it as a lower band, the transcripts lacking exon 9, and as a higher band, the transcripts containing exon 9. So exon 9 is expressed mainly in IFM and only weakly in TDT muscles. (D) Using exon 2 and exon 9 probes for TpnI, it can be seen that exon 3 only appears heavily expressed in thoracic muscles including always exon 9. RT-negative control lanes are not shown.

Differences in the expression profile of TpnT and TpnI transcripts in the Drosophila melanogaster musculature. Agarose 1.2% gel separations of RT-PCRs. Each transcript detected by PCR as a band is marked with an arrow labeled with the name of the amplified gene and its exon composition. (A) Expression pattern of mutually exclusive exon 10's detected in adult or embryonic RNA extractions with four exon 10A- or 10B-specific primers (10A1, 10A2, 10B1, and 10B2). (B) Patterns of expression in adult body parts (head, IFM, TDT, and abdomen). Using probes for TpnT exon 6 and 10A (10A2 probe) or 10B (10B1 probe), the exon 10 expression pattern is shown. Exon 10B is expressed in all muscles in the adult except in the IFM. Exon 10A is the only one expressed in the IFM, but it is also expressed in the TDT muscles. (C) Probes for the 3′ region of the TpnI gene were used in RT-PCRs (exons 8–9 and 8–10). Exon 9 uses the first two residues of exon 10 to generate a stop codon, so the exon 10 probe detects it as a lower band, the transcripts lacking exon 9, and as a higher band, the transcripts containing exon 9. So exon 9 is expressed mainly in IFM and only weakly in TDT muscles. (D) Using exon 2 and exon 9 probes for TpnI, it can be seen that exon 3 only appears heavily expressed in thoracic muscles including always exon 9. RT-negative control lanes are not shown.

Detection of exon 10A was particularly strong in the thorax, suggesting a specific role of the sequence encoded by this exon in the flight-related musculature (data not shown). Exon 10A- and 10B-specific probes detected, by RT-PCR, the classic 10B exon in the head, abdomen, and dissected TDT fibers ( fig. 5B). Exon 10A expression was found almost exclusively in the RT-PCR product from RNA extracted from IFM and TDT fibers although some residual expression was also detected in abdomens. Consequently, IFM TpnT contains sequences encoded only by exon 10A, while TDT muscles contain a mixture of transcripts with exon 10A or 10B. The same thoracic enrichment of exon 10A transcripts was obtained in D. subobscura and D. virilis.

TpnI alternatively spliced exons were detected using several probe combinations. Exon 9 is present in TDT and IFM transcripts ( fig. 5C), but while in IFMs it is the only one expressed, in TDT it is almost completely replaced by the exon 10–containing transcripts. Using probes for the complete coding region (exon 2 to exon 9/10) in the adult body parts showed that the alternatively spliced exon 3 is expressed in TDT/IFM, but is only present in transcripts that also incorporate exon 9 ( fig. 5D).

Expression Profile of the TpnT and TpnI Genes in A. mellifera

The same procedure has been carried out with the honeybee to discover if the same patterns of isoform abundance occur. The expression profile of the TpnT transcripts during A. mellifera development ( fig. 6A) shows that larval muscles contain a mixture of transcripts with alternative exons 3, 4, and 5′ or exons 3 and 4. The IFMs contain an isoform encoded by a transcript lacking all the 5′ region alternatively spliced exons. Other adult muscles contain a mixture of transcripts containing the exons 3, 4, 5, and 5′ or 3, 4, and 5. In relation to the 3′ half of the genes, exon 10B–containing transcripts were found in all muscles except in IFM, where it is only marginally detected (dorsoventral indirect flight muscle) or not at all (dorsolateral indirect flight muscle). Exon 10A–containing transcripts are adult thorax-specific, similar to the drosophilid results (see above).

Expression profile of TpnI and TpnT transcripts in Apis mellifera. Each transcript detected by PCR as a band is marked with an arrow labeled with the name of the amplified gene and its exon composition. Adjacent empty lanes to each PCR lane are negative controls (mRNA without reverse transcription). (A) Troponin T expression profile during development (larval L1 and L2 and pupal P1 and P2 stages) and in the thorax muscles (dorsoventral indirect flight muscles [DV-IFM], dorsolateral indirect flight muscles [DL-IFM], legs, and other thoracic muscles) of Apis. Apis TpnT shows a similar expression pattern to that described in Drosophila. (B) The TpnI expression pattern in the same sample is shown. The alternative exons, 9 and 10, both containing a proper stop codon and terminator signals, are expressed in different levels in all muscles and stages. Using probes from exon 4 and exon H1 of the Apis TpnI gene, we have detected an IFM-specific expression pattern for this TpnH isoform in hymenopterans. The four exon 6's identified are expressed in the TpnI transcripts (detected by inner PCR using internal exon 6 type-specific probes in a second PCR using the P2 fraction band as substrate DNA), except in the IFM-specific transcript where only 6b exons are detected. (C) The 3′ rapid amplification cDNA extension (RACE) experiments were performed with thorax, IFM, and larval samples to detect the transcription stops in the Apis TpnI gene. Transcripts representing eight different stop sites have been proportionally represented on the left, with their size indicated and different arrows styles (depending on the transcripts last exon, dotted line for exon 9, continuous line for exon 10, and broken line for exon H's) signaling the corresponding band in the RACE electrophoresis gels. Three of them (one transcript for each group) are the major ones used in the thoracic muscles (their lengths are shown inside circles), but the transcript containing the TpnH exons is absent in larval muscles. RACE-negative control lanes are shown (− lanes).

Expression profile of TpnI and TpnT transcripts in Apis mellifera. Each transcript detected by PCR as a band is marked with an arrow labeled with the name of the amplified gene and its exon composition. Adjacent empty lanes to each PCR lane are negative controls (mRNA without reverse transcription). (A) Troponin T expression profile during development (larval L1 and L2 and pupal P1 and P2 stages) and in the thorax muscles (dorsoventral indirect flight muscles [DV-IFM], dorsolateral indirect flight muscles [DL-IFM], legs, and other thoracic muscles) of Apis. Apis TpnT shows a similar expression pattern to that described in Drosophila. (B) The TpnI expression pattern in the same sample is shown. The alternative exons, 9 and 10, both containing a proper stop codon and terminator signals, are expressed in different levels in all muscles and stages. Using probes from exon 4 and exon H1 of the Apis TpnI gene, we have detected an IFM-specific expression pattern for this TpnH isoform in hymenopterans. The four exon 6's identified are expressed in the TpnI transcripts (detected by inner PCR using internal exon 6 type-specific probes in a second PCR using the P2 fraction band as substrate DNA), except in the IFM-specific transcript where only 6b exons are detected. (C) The 3′ rapid amplification cDNA extension (RACE) experiments were performed with thorax, IFM, and larval samples to detect the transcription stops in the Apis TpnI gene. Transcripts representing eight different stop sites have been proportionally represented on the left, with their size indicated and different arrows styles (depending on the transcripts last exon, dotted line for exon 9, continuous line for exon 10, and broken line for exon H's) signaling the corresponding band in the RACE electrophoresis gels. Three of them (one transcript for each group) are the major ones used in the thoracic muscles (their lengths are shown inside circles), but the transcript containing the TpnH exons is absent in larval muscles. RACE-negative control lanes are shown (− lanes).

The TpnI expression pattern detected using RT-PCR is shown in figure 6B. Probes for the region between exons 4 and 9 or exons 4 and 10 revealed that transcripts containing exon 9 or exon 10 are detected in the different developmental stages and in all muscles of the adult Apis musculature. Despite inherent limitations to these PCR experiments, exon 9–containing transcripts seem to appear later than those containing exon 10. Using probes for each specific exon 6 and exon 8 in an inner PCR of the previous amplified products, the four exon 6's in both exon 9– and exon 10–containing transcripts were detected. Although quantitative PCR was not carried out, the results indicate that maximum expression levels of exons 6a and 10 occur during larval stages, while those of 6b and 9 exons occur in adults.

Finally, RT-PCR using probes designed to amplify the region between exon 4 and exon H1 of the Apis TpnI gene showed, as expected, that a heavy Apis TpnI (troponin H) is expressed exclusively in the IFM ( fig. 6B). This band was cloned and tested for its exon 6 content, using an exon 8 probe in combination with a specific probe for each one of the four exon 6. Sequencing of 14 clones confirmed that 7 clones contained exon 6b1 and the other 7 clones contained exon 6b2. In addition, these TpnH transcripts include a shorter spliced version of exon 10 that could not be detected in earlier RT-PCR assays using the exon 10 probe. In order to locate the 3′ ends of the TpnI transcripts, we have performed rapid amplification cDNA extension experiments with three representative Apis RNA fractions ( fig. 6C). We found up to eight different transcript ends for the TpnI gene, but only three major ones occur in mRNA from the thoracic muscles and IFM samples. Two of them are 3′ to exon 9 and exon 10, produce the standard TpnI transcripts, and are also used in the larval muscles. The other end incorporates the initial part of exon 10 but extends into exons H1, H2, and H3, producing a TpnH isoform with a long alanine/proline-rich extension. We have also detected a transcript in which exon H1 does not use its acceptor-splicing site and this would produce a TnH isoform with a shorter alanine/proline-rich extension.


Genetic Basis, Diagnosis, and Management

Background: Congenital central hypoventilation syndrome (CCHS) is characterized by alveolar hypoventilation and autonomic dysregulation.

Purpose: (1) To demonstrate the importance of PHOX2B testing in diagnosing and treating patients with CCHS, (2) to summarize recent advances in understanding how mutations in the PHOX2B gene lead to the CCHS phenotype, and (3) to provide an update on recommendations for diagnosis and treatment of patients with CCHS.

Methods: Committee members were invited on the basis of their expertise in CCHS and asked to review the current state of the science by independently completing literature searches. Consensus on recommendations was reached by agreement among members of the Committee.

Results: A review of pertinent literature allowed for the development of a document that summarizes recent advances in understanding CCHS and expert interpretation of the evidence for management of affected patients.

Conclusions: A PHOX2B mutation is required to confirm the diagnosis of CCHS. Knowledge of the specific PHOX2B mutation aids in anticipating the CCHS phenotype severity. Parents of patients with CCHS should be tested for PHOX2B mutations. Maintaining a high index of suspicion in cases of unexplained alveolar hypoventilation will likely identify a higher incidence of milder cases of CCHS. Recommended management options aimed toward maximizing safety and optimizing neurocognitive outcome include: (1) biannual then annual in-hospital comprehensive evaluation with (i) physiologic studies during awake and asleep states to assess ventilatory needs during varying levels of activity and concentration, in all stages of sleep, with spontaneous breathing, and with artificial ventilation, and to assess ventilatory responsiveness to physiologic challenges while awake and asleep, (ii) 72-hour Holter monitoring, (iii) echocardiogram, (iv) evaluation of ANS dysregulation across all organ systems affected by the ANS, and (v) formal neurocognitive assessment (2) barium enema or manometry and/or full thickness rectal biopsy for patients with a history of constipation and (3) imaging for neural crest tumors in individuals at greatest risk based on PHOX2B mutation.

PHOX2B: The Disease-defining Gene for CCHS

PHOX2B Mutations in CCHS

PHOX2B Genotype/CCHS Phenotype

Mosaicism in a Subset of Parents with CCHS Children

Inheritance of CCHS and the PHOX2B Mutation

Mechanism by Which Mutations in PHOX2B Gene Result in CCHS Phenotype

A Model for Transitional and Translational Autonomic Medicine

In 1999 the American Thoracic Society published the first Statement on Congenital Central Hypoventilation Syndrome (CCHS) (1). Since then, the world of CCHS has exploded with (1) the discovery that the paired-like homeobox 2B (PHOX2B) gene is the disease-defining gene for CCHS (2–5) (2) identification of an autosomal dominant inheritance pattern (3, 5–7) (3) demonstration of a PHOX2B genotype–CCHS phenotype relationship pertinent to ventilatory dependence (3, 5), facial dysmorphology (8), cardiac asystoles (9), and Hirschsprung disease and neuroblastoma (6, 7) (4) identification of PHOX2B mutations in CCHS adults and older children (10–17) whose diagnosis was “missed” or not apparent in the neonatal period, infancy, and early childhood (5) documentation of mosaicism in 5 to 10% of parents of children with CCHS (3, 6) and (6) improved understanding of the specific mechanisms whereby PHOX2B results in the CCHS phenotype (6, 18–21). The purpose of a new ATS statement on CCHS is to aid the clinician in optimizing patient care that will be specifically tailored to knowledge of the individual PHOX2B genotype/mutation, and to offer genetic counseling. Finally, all management options will be addressed with the long-term goals of improving quality of life for PHOX2B mutation-confirmed individuals with CCHS and to gain a better understanding of the autonomic nervous system (ANS) in health and disease. These include (1) biannual then annual in-hospital evaluation with (i) physiologic studies during awake and asleep states to assess ventilatory needs during varying levels of activity and concentration, in all stages of sleep, with spontaneous breathing, and with artificial ventilation, and to assess ventilatory responsiveness to physiologic challenges while awake and asleep, (ii) 72-hour Holter monitoring, (iii) echocardiogram, (iv) assessment of ANS dysregulation across all organ systems affected by the ANS, and (v) formal neurocognitive assessment (2) barium enema or manometry and/or full thickness rectal biopsy for patients with a history of constipation and (3) imaging for neural crest tumors in individuals at greatest risk based on PHOX2B mutation.

Committee members were invited on the basis of their expertise in the care of patients with CCHS in terms of clinical care, clinical research, or basic science investigation. Representation was included from the pediatric and adult communities. The intent was for international representation. Committee members were asked to review the current state of the science by independently completing literature searches using Pub Med and OVID. Each committee member was asked to assess the identified literature, provide a critique of articles, and rate the importance of individual articles. The most highly ranked articles and the critiques were integrated into the working document. Consensus on the recommendations was reached among the members of the Committee.

To inform the practitioner, parent, caregiver, and health care provider that a PHOX2B mutation is requisite to confirmation of a diagnosis of CCHS.

To improve general knowledge regarding PHOX2B as the disease-defining gene for CCHS. The reader will learn that (i) approximately 90% of individuals with the CCHS phenotype are heterozygous for a polyalanine expansion repeat mutation in the PHOX2B gene, (ii) approximately 10% of individuals with CCHS are heterozygous for a missense, nonsense, or frameshift mutation in the PHOX2B gene, (iii) other non-CCHS diagnoses should be sought if a PHOX2B mutation is not found.

To introduce the opportunity to anticipate the CCHS phenotype based on the PHOX2B genotype/mutation.

To educate clinicians that CCHS is no longer diagnosed exclusively in the newborn period, as it is now described among toddlers, children, and adults.

To focus on the autosomal dominant inheritance pattern of the PHOX2B mutation in CCHS, the finding of mosaicism in 5 to 10% of parents, and the importance of testing both parents of each subject with CCHS.

To improve understanding of the specific mechanisms whereby PHOX2B results in the CCHS phenotype.

To update information regarding available treatment and home health care options.

To recognize that CCHS is a model for translational and transitional autonomic medicine. In addition to using the PHOX2B genetic mutation to optimize patient management, there will be a need for clinicians to continue to care for these special patients as they mature into adulthood.

CCHS was first described in 1970 by Robert Mellins and colleagues (22). Despite a multitude of case reports, large series were not published until 1992 (23). The 1999 ATS Statement on CCHS estimated “roughly 160 to 180 living children with CCHS worldwide” but advised that these numbers “are considered to be an underestimate” (1). In 2009, the collective laboratories from the United States, France, Italy, Japan, Germany, Taiwan, China, The Netherlands, Chile, the UK, and Australia have now diagnosed nearly 1,000 cases with PHOX2B mutation-confirmed CCHS. Even now, this is recognized to be an underestimate, as individuals with the milder phenotype are underdiagnosed. Although CCHS is characteristically diagnosed during the newborn period, recent reports indicate that individuals can be diagnosed in childhood (5, 6, 17, 24–26) and adulthood (10–17, 26), depending upon the PHOX2B genotype and the intellectual inquisitiveness of the patient, family, and medical team. Regardless of age at presentation, individuals with CCHS will be clinically diagnosed in the absence of primary lung, cardiac, or neuromuscular disease or an identifiable brainstem lesion that might account for the entire phenotype inclusive of the autonomic nervous system dysregulation (ANSD). Individuals with CCHS characteristically have diminutive tidal volumes and monotonous respiratory rates awake and asleep (1), although the more profound alveolar hypoventilation occurs primarily sleep. As a result of the hypoventilation, these individuals will become hypoxemic and hypercarbic but typically lack the responsiveness to these endogenous challenges in terms of ventilation and arousal during sleep, and they lack the perception of asphyxia during wakefulness with and without exertion (1). Conditions associated with CCHS reflecting anatomic ANSD include Hirschsprung disease (HSCR) and tumors of neural crest origin in addition to a spectrum of symptoms compatible with physiologic ANSD, including diminished pupillary light response, esophageal dysmotility, breath-holding spells, reduced basal body temperature, sporadic profuse sweating, lack of perception to dyspnea, altered perception of anxiety, and lack of physiologic responsiveness to the challenges of exercise and environmental stressors (1, 23, 27–39). CCHS is a lifelong disease, which raises key questions including: (1) will the phenotype change with advancing age based on the PHOX2B mutation, age at diagnosis, and adequacy of management? and (2) will intervention strategies be effective considering the nature of the mutation and its timing in terms of embryologic development? As the aim of this Statement is not to provide an exhaustive review of CCHS, the reader is referred to recent reviews (www.genereviews.org and References 40 and 41).

Hints toward the familiality of CCHS emerged between the 1980s and 2001. Familial recurrence data include one report each of affected monozygotic female twins (42), sisters (43), male-female siblings (23, 44), and male-female half siblings (45) with CCHS. In the pre-PHOX2B/CCHS era, five women diagnosed with CCHS in their own childhoods gave birth to two infants with definite CCHS, one with likely CCHS confounded by severe immaturity and bronchopulmonary dysplasia, and one with later-onset CCHS (46, 47). A report of a child with CCHS born to a woman who had neuroblastoma as an infant (48) added to the premise of a transmitted genetic component in the phenotypic spectrum of ANSD and CCHS. Furthermore, ANSD was studied in a case-control family design (27, 28), which provided important confirmatory evidence for a genetic basis to CCHS and is therefore regarded as the most severe manifestation of a general ANSD (1, 27, 44), although the role of PHOX2B in the broader category of ANSD remains unknown.

Most early studies, undertaken in pursuit of the genetic basis for CCHS, were restricted to genes known to be related to HSCR. Twenty patients were reported to have protein-altering mutations in receptor tyrosine kinase (RET) (49–53), glial cell–derived neurotrophic factor (GDNF) (49), endothelin signaling pathway 3 (EDN3) (52, 54), brain-derived neurotrophic factor (BDNF) (55), human aschaete-scute homolog gene (HASH1) (4, 56), paired-like homeobox gene 2A (PHOX2A) (4), GFRA1 (4), bone morphogenic protein 2 (BMP2) (3), and endothelin converting enzyme 1 (ECE1) (3). Three other reports indicate an absence of RET (57) and RNX mutations (58, 59).

In 2003, PHOX2B was found to be the disease-defining gene for CCHS (2, 3). PHOX2B encodes a highly conserved homeodomain transcription factor known to play a key role in the development of ANS reflex circuits in mice (60, 61). PHOX2B contains a repeat sequence of 20 alanines in exon 3, which Amiel and colleagues reported to contain in-frame duplications of 15 to 27 nucleotides, leading to expansion of the repeat tract to 25 to 29 alanines on the affected allele in 18 of 29 (62%) French CCHS cases (2). These expansions appeared to be de novo insofar as they were not present in eight sets of parents of the CCHS cases. Two of 29 (7%) CCHS cases had frameshift mutations. Collectively, 69% of the 29 French patients were heterozygous for a PHOX2B mutation, but none of the controls had PHOX2B mutations. Amiel and colleagues (2) also demonstrated PHOX2B expression in early human embryos in both central autonomic neuron circuits and in peripheral neural crest derivatives.

Concurrent to the French studies, Weese-Mayer and colleagues (3) focused on genes involved in the early embryology of the ANS (mammalian aschaete-scute homolog-1 [MASH1], BMP2, engrailed-1 [EN1], TLX3, ECE1, endothelin-1 [EDN1], and PHOX2A). Although no novel disease-causing mutations were found in any of these genes in a cohort of 67 CCHS cases, Weese-Mayer and colleagues (3) identified heterozygous PHOX2B exon 3 polyalanine repeat expansions of 25 to 33 repeats in 65 of 67 (97%) children with the CCHS phenotype. Of the two remaining CCHS cases, a nonsense mutation (premature stop codon) in PHOX2B was identified in one patient and the other was later found to have a polyalanine repeat expansion in PHOX2B after a sample mix-up at the lab of origin was resolved (7). Collectively, Weese-Mayer and colleagues (3) identified mutations in exon 3 of the PHOX2B gene in 100% of the 67 children with the CCHS phenotype, indicating that PHOX2B is the disease-defining gene in CCHS. None of the PHOX2B expansion mutations were present in 67 gender/ethnicity-matched controls. This study also noted (1) an association between polyalanine expansion length and severity of autonomic dysfunction (2) mosaicism in 4 of 97 parents of CCHS cases, suggesting that not all PHOX2B mutations occur de novo (3) autosomal dominant inheritance of the PHOX2B mutation and the CCHS phenotype from CCHS cases and (4) autosomal dominant inheritance of the PHOX2B mutation from mosaic parents. Furthermore, these authors (3) established the first clinically available assay for the diagnosis of CCHS using a simple and accurate method for detecting and sizing the repeat sequence associated with the polyalanine tract expansion (patented Rush University Medical Center, Chicago, IL patent donated to charitable trust and proceeds from PHOX2B Screening Test support CCHS research), which could also be used for prenatal diagnosis, family testing, and diagnosis of individuals with relevant symptoms.

Subsequent to the above studies, PHOX2B polyalanine repeat expansions were found in 4 (40%) and a PHOX2B insertion frameshift mutation in 1 (10%) of 10 CCHS cases in Japan (4). The expansion was shown to be de novo in 2 cases. Sasaki and colleagues (4) used the same methodology as the French (2) and also underdetected PHOX2B expansion cases as reported in 2005 (62). In 2004, Matera and colleagues (5) identified heterozygous PHOX2B polyalanine expansion mutations of 25 to 33 repeats in 21 (88%) and heterozygous frameshift mutations in 2 (8%) of 24 CCHS cases from Italy, Germany, and The Netherlands. This study confirmed the correlation between the size of the PHOX2B expanded allele and the severity of the respiratory phenotype and associated symptoms (3). Matera and colleagues also demonstrated that in standard polymerase chain reactions the CCHS-associated expanded allele, especially those with 30 to 33 alanines, can remain undetected due to the GC-rich polyalanine region of PHOX2B thus, the PHOX2B mutation rate may be underestimated as a result of the amplification-induced allele dropout. Using assays designed to amplify GC-rich regions, Trang and colleagues (63) (re)analyzed 34 of the French patients and identified a PHOX2B mutation in 91% of the cases, and Trochet and colleagues (6) found PHOX2B mutations in 93% of 174 subjects with CCHS from multiple nationalities, including 7 of 9 “mutation-negative” patients reported by Amiel and colleagues in 2003 (2). Berry-Kravis and colleagues (7) and Weese-Mayer and colleagues (64) reported PHOX2B mutations in 184 subjects (in 2006) and collectively more than 350 subjects (in 2008) with CCHS, respectively (100% sensitivity and specificity of detection), in a cohort primarily from the United States, with 10% of patients from abroad.

As previously summarized (40, 41) (www.genereviews.org), the range for the number of repeats in the PHOX2B polyalanine expansion on the affected allele in patients with CCHS is 24 to 33 (2–7, 13, 15–17, 25, 26, 65–70). Polyalanine repeat expansion mutations (PARMs) were not found in 482 controls from the above-cited publications nor among 1520 healthy individuals in Taiwan (67). In-frame contraction variants (with 7, 13, 14, or 15 repeats in the polyalanine repeat tract) have been reported in three CCHS cases (3, 7, 71) who harbor an additional polyalanine expansion repeat mutation or nonpolyalanine repeat mutation (NPARM) but are also found in approximately 3% of seemingly normal controls (2, 3, 5, 72, 73), CCHS parents (3, 16, 71), and a small subset of patients with vague symptoms suggestive of autonomic dysregulation and/or sporadic hypoventilation, or apparent life-threatening events but not the constellation of symptoms characteristic of CCHS (71). A PHOX2B polyalanine repeat expansion mutation segregating with disease was observed in 9 of 16 patients with RET, GDNF, BDNF, HASH1, and GFRA1 coincidental mutations, thus indicating that PHOX2B is the disease-defining gene in these children.

A mutation in the PHOX2B gene is requisite to a diagnosis of CCHS. Over 90% of CCHS cases will be heterozygous for an in-frame PARM coding for 24 to 33 alanines in the mutated protein and producing genotypes of 20/24 to 20/33 (the normal genotype would be referred to as 20/20). The remaining approximately 10% of patients with a classical CCHS phenotype will be heterozygous for an NPARM (74) (including missense, nonsense, and frameshift) in the PHOX2B gene. The 20/25, 20/26, and 20/27 genotypes are the most common, although growing numbers of even the less-common mutations are being identified monthly (see a histogram of all published data as well as all current data from the authors as of late 2009 in Figure 1 ).

Figure 1. Number of PHOX2B polyalanine repeat mutations by genotype. These data represent all published literature as well as the provided current data from the Statement authors for the polyalanine repeat mutations (PARMs) and the nonpolyalanine repeat mutations (NPARMs). The most common genotypes are 20/25, 20/26, and 20/27. Adapted by permission from Reference 41.

NPARMs (74) have been reported in association with CCHS by groups in the United States (3, 7, 20, 41, 64), Italy (5, 16, 18), Japan (4), France (2, 6, 13, 75), Germany (50, 69, 76), Australia (77), The Netherlands (78), China (79), and Taiwan (67, 80). Thus far, 76 individuals with CCHS and NPARMs in PHOX2B have been described worldwide, and mutations include predominantly frameshift mutations (59/76, 78%), but also nonsense (3/76, 4%), missense (12/76, 16%), and missense with stop codon alteration (2/76, 3%) (see Figure 2 for a schematic of all published data as well as all current data from the Statement authors). The majority of CCHS-associated NPARMs are found at the end of exon 2 or in exon 3 ( Figure 2 ).

Figure 2. Schematic for the PHOX2B gene with location of all CCHS-associated mutations described to date in PHOX2B. All polyalanine repeat mutations (PARMs) are located within the second polyalanine stretch of exon 3. Nearly all thus far identified NPARMs are found at the 3′ end of exon 2 or in exon 3. These data represent all published literature and current data provided by the Statement authors (41). Reproduced by permission from Reference 41.

The majority of NPARMs occur de novo and produce very severe phenotypes with HSCR and extensive gut involvement, need for continuous ventilatory support, and increased tumor risk in those over 1 year of age (6, 7). Thus, the presence of extensive HSCR and a CCHS phenotype is a strong predictor of a PHOX2B NPARM. Recurrent 38 and 35 base pair deletions, causing frameshift from the polyalanine repeat throughout the protein, produce severe disease and have been identified by several groups in different countries. A minority of NPARMs are associated with a high incidence of HSCR but a milder physiologic CCHS phenotype, and incomplete penetrance in at least 3 families (7). A few similarly located frameshift mutations (618delC, 577delG) have been inherited and are variably penetrant in families (5, 7), suggesting that −1 frameshifts in this area may produce a milder cellular deficit than other frameshift mutations. The c.422G > A and c.428A > G mutations, leading to p.R141Q and p.Q143R, respectively, have also been found in several unrelated cases of CCHS and, together with the c.299G > T (p.R100L) mutation (20), are the only missense mutations yet identified in CCHS. The c.419C > A (p.A140E) has recently been reported in later-onset CCHS, both isolated and associated with HSCR (13, 81).

Despite identification that PHOX2B is the disease-defining gene for CCHS in 2003, journals continue to publish research without (1) confirmation that all subjects with “CCHS” have PHOX2B mutations, (2) a distinction in data analysis between subjects with PHOX2B mutation-confirmed CCHS and children with other causes of hypoventilation, and (3) analysis of data in a PHOX2B genotype/CCHS phenotype format. Owing to the crucial role of PHOX2B in the development of the ANS, it is pertinent to hypothesize a relationship between PHOX2B genotype and the following aspects of the CCHS phenotype.

There is a relationship between the genotype for PARMs and the need for continuous ventilatory dependence (3, 5, 7, 65). Specifically, individuals with the 20/25 genotype rarely require 24-hour per day ventilatory support individuals with the 20/26 genotype have variable awake needs depending upon the level of activity and individuals with genotypes from 20/27 to 20/33 typically require continuous ventilatory support. Later-onset cases with the 20/24 or 20/25 genotype (10, 11, 17) have the mildest hypoventilation, presenting primarily after exposure to respiratory depressants or severe respiratory infection, and are managed with nocturnal ventilatory support only. In contrast to the PARMs, most individuals with NPARMs require continuous ventilatory support (7) ( Figure 3 ).

Figure 3. Rate of continuous ventilatory dependence, Hirschsprung disease, and tumors of the neural crest in congenital central hypoventilation syndrome (CCHS) cases with polyalanine repeat expansion mutations (PARMs) in PHOX2B compared with CCHS cases with nonpolyalanine repeat expansion mutations (NPARMs) in PHOX2B. CCHS cases included in this figure were compiled from all known cases reported in the literature, including reports from groups in the United States, Italy, France, Japan, Germany, Taiwan, China, Australia, and The Netherlands as well as current information reported by the Statement authors, where adequate clinical information was available. Neural crest tumor data was derived from cases in which information was available and the child had survived at least the first year of life. All PARM cases with tumors had large (29–33 repeat) expansion mutations. Adapted by permission from Reference 41.

Long recognized to occur among 20% of cases of CCHS, HSCR is more clearly prevalent among cases of the NPARMs than the PARMs. Specifically, HSCR is reported in 87 to 100% of NPARMs in contrast to 13 to 20% of PARMs (6, 7, 65) (see Figure 3 for a histogram of all adequately detailed published data as well as all provided current data from the ATS Statement authors). Among the PARMs, there are no reports of HSCR occurring in subjects with the 20/25 genotype, and only rarely with the 20/26 genotype. A high occurrence of HSCR in individuals with the 20/27 genotype was reported in one cohort (6) but was not yet definitively confirmed in others. Recent studies further suggest that the RET gene may have a pivotal role as a modifier gene for the HSCR phenotype in patients with CCHS (53, 82).

Tumors of neural crest origin occur more frequently among individuals with NPARMs (50%) than among those with PARMs (1%) (6, 7, 65) ( Figure 3 ) (all neuroblastomas). However, among PARMs, only subjects with the 20/29 and 20/33 genotypes (2, 3, 6, 7) have been identified to have tumors of neural crest origin (ganglioneuromas and ganglioneuroblastomas) thus far.

A recent report by Gronli and colleagues (9) identified a correlation between the most common PARMs (genotypes 20/25–20/27) and length of R-R intervals on Holter monitoring. Specifically, none of the children with the 20/25 genotype had sinus pauses of 3 seconds or longer. However, 19% of individuals with the 20/26 genotype and 83% of individuals with the 20/27 genotype had pauses of 3 seconds or longer. Similarly, cardiac pacemakers were implanted among 0% of the subjects with the 20/25 genotype, 25% of subjects with the 20/26 genotype, and 67% of subjects with the 20/27 genotype. Among the cases with the 20/26 and 20/27 genotypes who did not receive a cardiac pacemaker, two died suddenly and one had severe neurocognitive compromise. Further, one adult with the 20/25 genotype, diagnosed with CCHS in adulthood, had documented pauses of 8 seconds and longer (11) and the ATS Statement authors have been alerted to another later-onset adult with the 20/25 genotype and prolonged asystoles on Holter recording. These findings raise concern that individuals with the 20/25 genotype may be unaffected during childhood but, if not adequately managed or promptly diagnosed, may experience prolonged asystoles in adulthood. Risk to individuals with NPARMs remains unascertained.

Weese-Mayer and colleagues and Patwari and colleagues (3, 83) demonstrated that an increased number of polyalanine repeats was associated with an increased number of symptoms of ANS dysregulation ( Figure 4 ). Though these measures of ANSD were ascertained from review of medical records, scripted questionnaires, and physiologic assessment, they did not include specific tests to assess autonomic function. However, physicians and parents should expect more symptoms of ANSD among subjects with genotypes 20/27 to 20/33.

Figure 4. Number of autonomic nervous system dysregulation (ANSD) symptoms in congenital central hypoventilation syndrome (CCHS) cases versus PHOX2B genotype among 65 children with a polyalanine repeat expansion mutation (PARM). The number of symptoms of ANSD increases with the number of alanines in the PARM. Many subjects had identical numbers for ANSD symptoms and genotype, therefore the figure gives the illusion of fewer data points than expected for the cohort size. Measured genotype analysis (consisting of comparing the genotypic means by ANOVA) revealed a significant association between PHOX2B polyalanine repeat mutation length and number of symptoms of ANS dysregulation (F = 2.93, df = 5, P = 0.021). Adapted by permission from Reference 3.

A report by Todd and colleagues (8) described a characteristic facies among children between the ages of 2 years and early adulthood that have CCHS and primarily among individuals with PARMs. The faces of subjects with CCHS were generally shorter and flatter, and typically showed an inferior inflection of the lateral 1/3 of the upper vermilion border (lip trait). Though not dysmorphic, the face is short relative to its width, resulting in the characteristic box-shaped face observed in CCHS. Eighty-six percent of the CCHS cases and 82% of controls were correctly predicted using five variables to characterize facies (upper lip height, biocular width, upper facial height, nasal tip protrusion, and the lip trait). A limited number of cases with higher numbers of repeats (only five cases with 30–33 repeats) may have precluded the identification of a significant correlation between the number of polyalanine repeats and measures of the CCHS facial phenotype.

Todd and colleagues (84) assessed dermatoglyphic pattern type frequency, left/right symmetry and genotype/phenotype correlation in CCHS. Dermatoglyphic pattern type frequencies were altered in cases of CCHS versus controls: increase of arches in females and ulnar loops in males, with the largest differences for the left hand and for individuals with both CCHS and HSCR, was reported. Dissimilarity scores between the CCHS and CCHS/HSCR cases, and between all female and all male cases were not significantly different. No significant association was found between the number of polyalanine repeats in the PHOX2B genotypic category and dermatoglyphic pattern frequencies in the CCHS study groups.

The term “congenital” as used historically in CCHS, connoted presentation in the newborn period. However, patients presenting outside of the newborn period with later-onset (LO)–CCHS have been described previously (5, 6, 10–17, 24–26, 85). In the context of (1) increased awareness of CCHS (2) the discovery that PHOX2B is the disease-defining gene for CCHS and 3) the availability of clinical diagnostic testing for PHOX2B mutations, an increase in diagnosis of LO–CCHS with presentation in later infancy, childhood, and adulthood is anticipated.

LO-CCHS reflects the variable penetrance of the PHOX2B mutations with the genotypes 20/24 and 20/25 or rarely an NPARM that may require an environmental cofactor to elicit the phenotype. Careful review of the medical history for individuals “presenting” with alveolar hypoventilation after the newborn period often demonstrates signs and symptoms compatible with prior hypoventilation and other disorders of autonomic regulation from the newborn period. The diagnosis of LO–CCHS should be considered in cases of centrally mediated alveolar hypoventilation and/or cyanosis or seizures noted after (1) administration of anesthetics or CNS depressants, (2) recent severe pulmonic infection, or (3) treatment of obstructive sleep apnea. With a heightened clinical suspicion of LO–CCHS, the physician can expedite the diagnosis by promptly testing for a PHOX2B mutation, thereby averting potentially life-threatening decompensation as well as risk for neurocognitive compromise. Evaluation of later presentation cases requires a careful history with attention to past exposure to anesthesia or sedation, delayed “recovery” from a severe respiratory illness, and unexplained seizures or neurocognitive impairment. Also, review of digital frontal and lateral photographs (to evaluate for facies consistent with CCHS adult males often have a moustache to conceal the “lip trait”), any electrocardiographic documentation of prolonged sinus pauses (ideally via 72-h Holter monitoring), any physiologic evaluations documenting ventilation while awake and while asleep (for hypercarbia and/or hypoxemia), a hematocrit and reticulocyte count (for polycythemia and response to hypoxemia), a bicarbonate level (for signs of compensated respiratory acidosis), or chest x-ray, echocardiogram, or electrocardiogram (for signs of right chamber enlargement or pulmonary hypertension) should be completed. In cases of constipation, a barium enema or manometry may be considered to exclude short segment HSCR. Patients diagnosed after the neonatal period would be termed LO–CCHS and can be distinguished from other syndromes of mild alveolar hypoventilation by the presence of a PHOX2B mutation.

The 20/24 genotype is likely underdiagnosed because of the subtle hypoventilation and potential need for environmental cofactors (17) or the homozygous condition to manifest the CCHS phenotype (66). Molecular analyses of PHOX2B in cohorts of individuals presenting with profound hypoventilation after anesthesia, sedation, or respiratory illness may identify additional patients with the 20/24 and 20/25 genotype. In so doing, other yet-unidentified environmental or genetic factors that might impact the variable penetrance of the 20/24 and 20/25 mutations, as well as the most recently described novel missense mutation (81), may be determined.

It is essential that practitioners distinguish LO–CCHS from rapid-onset obesity with hypothalamic dysfunction, hypoventilation, and autonomic dysregulation (ROHHAD) (86), a rare disorder first described in 1965 (87). Originally termed late-onset central hypoventilation syndrome with hypothalamic dysfunction (88), it was renamed in 2007 (86) to alert the practitioner to the typical sequence of presenting symptoms. In that same publication, Ize-Ludlow and colleagues clarified that ROHHAD is a distinctly different syndrome from CCHS as demonstrated by careful history and PHOX2B testing. Although fewer than 55 children have been described in the literature with this disorder (86, 88, 89), it is essential to recognize the phenotype as distinct from CCHS. Children with ROHHAD typically present between the ages of 1.5 and 7 years with rapid onset obesity (20–40 pound gain over 4–6 mo), followed by the recognition of other hypothalamic disorders including water imbalance, elevated prolactin levels, altered onset of puberty, and more. Nearly half of the children will experience a cardiorespiratory arrest after an intercurrent viral infection, then obstructive sleep apnea and hypoventilation will be noted. At variable times thereafter, symptoms of autonomic dysregulation including low body temperature, cold hands and feet, severe bradycardia, decreased pain perception, among others, will become apparent. At any point in the disease manifestation as many as 40% of children will demonstrate a tumor of neural crest origin, often associated with scoliosis. Behavioral disorders, strabismus, and abnormal pupillary responses are all reported in ROHHAD. These children can often be supported with mask ventilation only at night, but a subset of them will require 24-hour per day ventilation via tracheostomy. Although ongoing studies into the possible causes of ROHHAD are underway, the specific cause for this disorder has not been identified. As there is no genetic testing available for this disorder, the diagnosis of ROHHAD is based on the clinical presentation, the related clinical features, and documented absence of other potentially confounding diagnoses, including ruling out CCHS with clinically available PHOX2B testing (documenting an absence of PARMs and NPARMs).

Most expansion mutations occur de novo in CCHS, but 5 to 10% are inherited from a mosaic typically unaffected parent. A distinction is needed between germline inheritance and somatic occurrence of the PHOX2B mutation. Incomplete penetrance has been demonstrated when certain PHOX2B mutations are present in all the cells (including the reproductive cells of the germline) of individuals known to be unaffected. These latter PHOX2B mutations (20/24, 20/25, and a few NPARMs), although asymptomatic in some individuals, may be characterized by milder or variable phenotypic effects in the children affected with CCHS or other family members (3, 5, 17), respectively. In contrast, somatic mosaicism, due to postzygotic mutations, has been reported among a subset of parents of typical CCHS cases carrying PHOX2B polyalanine (PA) alleles larger than 25 repeats and NPARMs (3, 6, 7, 16).

Somatic mosaicism for a PARM was first reported in 2003 by Weese-Mayer and colleagues (3) in 4 parents out of 54 available families (7.4%). In 2005 Trochet and colleagues (6) identified somatic mosaicism in 1 parent of each of 10 CCHS patients, confirming that roughly 10% of children with CCHS will inherit the mutation from a mosaic parent. In both studies, mosaicism was detected in DNA extracted from parents' peripheral leukocytes by observing a lighter signal from the expanded allele than from the normal allele, in contrast to the pattern seen in subjects with CCHS.

A quantitative estimate of the somatic mosaicism in unaffected parents has recently been assessed in two other studies. DNA amplification products from asymptomatic carriers of alanine expansions with genotypes ranging from 20/25 to 20/31 were loaded on a DNA automated sequencer and expanded, and normal alleles were visualized ( Figure 5 ) as output peaks whose underlying area was directly proportional to their respective amounts. Although the mutant peak was expected to represent 50% of the PHOX2B alleles in individuals who inherited the mutation, it was found to range from 9 to 35% in DNA from parental leukocytes this percentage was confirmed in studies of fibroblast and saliva DNA from a subset of these mosaic parents (13). Likewise, in another study, mosaic individuals were identified as “outliers,” with less signal in the peak corresponding to the expanded allele (16). Whereas somatic mosaicism for PARMs larger than 20/25 was demonstrated in these studies, none of the rare seemingly asymptomatic 20/25 carriers were found to be mosaic, confirming that in these cases lack of the disease phenotype can be ascribed to reduced penetrance of a germline mutation (13, 16). Taken together, these data support the hypothesis that germline PARMs larger than the 20/25 genotype are fully penetrant, and asymptomatic carriers may only be found in association with significant degrees of somatic mosaicism. The CCHS phenotype has not been associated with any degree of somatic mosaicism thus far, suggesting a germline origin for most PARMs in affected CCHS patients.

Figure 5. Differential amounts of wild-type and expanded PHOX2B alleles in PA mutation carriers. Blue peaks represent the PHOX2B alleles carried by a patient with congenital central hypoventilation syndrome (CCHS) (upper) and an asymptomatic parent (bottom). Although positions in the x-axis correspond to their lengths, as indicated underneath, peak height is directly proportional to their amount. In this light, the individual below presents a somatic mosaicism for the 26 Ala allele (genotype 20/26), which in fact correspond to much less than half with respect to the 20 Ala wild-type allele (normal genotype 20/20).

Detection of the same PHOX2B mutation in parent-child pairs and observation of somatic mosaicism in some unaffected parents for the mutation observed in their affected child clearly established an autosomal dominant inheritance pattern for CCHS (3, 6). Most parents of affected children with CCHS do not carry a mutation at all, indicating a high de novo mutation rate in affected individuals. The 20/24 and 20/25 genotype PARMs and some of the NPARMs may be found in the germline of asymptomatic parents of children with CCHS and even other family members, suggesting these mutations are inherited as dominant with incomplete penetrance (5, 7, 10, 11, 13, 17, 26). Family members who carry such mutations but do not have CCHS may show other ANSD phenotypes, including HSCR or neuroblastoma (7, 48), or may be presymptomatic, presenting in later childhood or adulthood.

Genetic counseling is crucial for individuals diagnosed with CCHS, their parents and, in some cases, specific family members. For all affected individuals with CCHS, there is a 50% chance of transmitting the mutation, and therefore the disease phenotype, to each offspring. If an unaffected parent is found to be mosaic for a PHOX2B mutation (usually identified because of an affected child), there will be up to a 50% chance of recurrence in any subsequent child. Mosaic individuals always can be assumed to have a new mutation (the mutation cannot be inherited in mosaic fashion) and therefore, only children of these individuals (not other family members) would be at risk to have the mutation. If unaffected parents do carry a germline mutation (i.e., a 20/25 genotype PARM) there may be numerous other family members who can carry the same mutation without having obvious symptoms. In this case, genetic testing is indicated for all persons in position in the pedigree to inherit the mutation, which can often be traced back until the individual in whom the mutation originated is identified. To assess recurrence risk in a family, both parents of children with CCHS should have the PHOX2B Screening Test done to rule out mosaicism (90) (PARMs of genotypes 20/26 to 20/33 and severe NPARMs) or a nonpenetrant carrier state (genotypes 20/24 and 20/25 PARMs and mild NPARMs). Prenatal testing is available and can be performed for individuals with or without CCHS who are known germline mutation carriers or recognized as somatic mosaics. Despite negative testing of parents of a CCHS child, germline mosaicism cannot be ruled out and prenatal testing for subsequent pregnancies should be considered. Prenatal testing can allow parents optimal information with which to make an informed decision with a range of possibilities from elective abortion to a fully prepared delivery room to optimize the baby's chance for a smooth transition to extrauterine life.

Polyalanine tracts, predicted in roughly 500 human proteins, are preferentially found in transcription factors and are regarded as flexible spacer elements essential to conformation, protein–protein interactions, and/or DNA binding. PA tracts coded by human genes, other than PHOX2B, have already been found expanded in association with at least nine different congenital disorders, including mental retardation and malformations of the brain, digits, and midline structures (91). In this light, PA expansions are members of a broader category of trinucleotide repeat–associated disorders that includes also polyglutamine (PQ) expansions. Unlike PQ tracts, PA stretches are generally stable (do not change size when passed on from one generation to another), are usually coded by imperfect trinucleotide repeats (alanine can be coded by four different DNA triplets) and, with the exception of rare contractions, are not present as polymorphic tracts in the human population (most wild-type genes all have the same number of alanines). These observations have suggested an unequal allelic homologous recombination (crossover) during meiosis and/or mitosis as the most attractive disease-causing mechanism for poly-A tract expansions (91).

However, in mosaic individuals, only two alleles (wild-type and expanded alleles), instead of the three alleles (wild-type, contracted, and expanded) expected after occurrence of a somatic event of unequal crossing-over, have been reported, demonstrating that an alternative mutational mechanism should be considered to explain the origin of these trinucleotide repeat expansions (16, 92). Indeed, by reasoning that imperfect trinucleotide repeat sequences, typical of PA tracts, would reduce the ability of the repeats to form misaligned structures, replication slippage has been proposed as a more plausible mechanism than unequal crossing over for the generation of PA expansions (93).

This view has recently been revised after observing four families informative for PHOX2B markers whose segregation was compatible with the occurrence of unequal sister chromatid exchange. It is probable that de novo expansion of PA repeats in CCHS results mainly from this sort of chromosomal event either during gametogenesis or in postzygotic somatic cells (94).

A paternal origin of the gametes transmitting expansions has been reported in six informative de novo CCHS trios (94), whereas in a larger cohort of 20 trios, 13 mutations occurred on the paternal and 7 on the maternal chromosomes. Thus, occurrence of PA repeat expansions may be independent from processes specific to sperm or oocyte development, or it may be that there is a weak gender bias that would require analysis of a larger sample of parent-child trios.

As a tissue-specific transcription factor, PHOX2B is responsible for the expression regulation of a series of target genes involved in the development of the ANS. The finding that PHOX2B binds directly to the regulatory regions of the dopamine-β-hydroxylase (DβH), PHOX2A, and TLX-2 genes has allowed application of a functional approach to disclose the molecular mechanisms underlying CCHS pathogenesis. To this end, PHOX2B mutations have been tested for potential disruption of the normal function of the protein with respect to (i) transactivation of different target promoters, (ii) DNA binding, (iii) aggregate formation, and (iv) subcellular localization. Distinct CCHS pathogenetic mechanisms for PARMs and NPARMs in PHOX2B have thus been postulated. In addition, the cellular response to PHOX2B polyalanine expansions has been investigated to determine whether there exist cellular mechanisms that could be targeted to limit the cytotoxicity of these mutations.

To investigate how PHOX2B PA expansions can induce CCHS pathogenesis, the ability of expression constructs containing PA mutations to regulate the transcription of known target genes has been compared, in two different laboratories, to a wild-type PHOX2B construct. In particular, as exemplified in Figure 6A for the DβH target gene, when mutant PHOX2B constructs were cotransfected with the DβH and PHOX2A regulatory regions connected to the Luciferase gene, a strict inverse correlation between the induced Luciferase activity and the length of the PA tract was identified. This suggested that the transcriptional regulation of these two genes is directly dependent on the correct structure of the PHOX2B domain, including the 20-alanine tract and that longer tracts increasingly disrupt transcription (6, 18). Finally, a significant reduction of the transactivating activity of PHOX2B constructs bearing different PA contractions on the DβH promoter has also been observed in the same reporter assay (72). Unfortunately, this observation could not be replicated in another cell recipient (6) suggesting the need for additional investigation before assessing whether, despite lack of any phenotypic effect, PA contractions result in some disruption of PHOX2B function.

Figure 6. Effect of PHOX2B mutations on transactivation of the DβH regulatory region. The transcriptional activity obtained by cotransfecting the PHOX2B expression constructs reported on the left of each diagram with a construct containing the DβH promoter cloned upstream of the Luciferase reporter gene is shown for both (A) poly-Ala expanded tracts and (B) frameshift mutations in terms of relative Luciferase activity. Adapted by permission from Reference 18.

Fluorescence microscopy of COS-7 cells expressing PHOX2B proteins fused to a green fluorescent molecule has shown that the wild-type PHOX2B protein is present almost exclusively in the nucleus. However, increasing length of the PA repeat does induce an increasing percentage of PHOX2B protein within cells to mislocalize to the cytoplasm ( Figure 7 ) (18). As such, similar experiments performed in HeLa cells induced formation of PHOX2B polyalanine aggregates, although in different amounts compared with that observed in COS-7 cells (6), suggesting that mislocalization of the mutant protein is a common pathogenetic mechanism leading to impaired transcriptional activity of mutant PHOX2B containing aggregation-prone expanded PA tracts.

Figure 7. Subcellular localization of proteins bearing different PHOX2B defects. In the panels on the right, three possible patterns of PHOX2B cellular localization are reported: (N) nuclear localization only (N+C) both nuclear and cytoplasmic localization, either diffuse or with formation of nuclear and/or cytoplasmic aggregates and (C) cytoplasmic localization only. Subcellular distribution of PHOX2B-GFP proteins, after cell-transfection with mutant constructs carrying different PHOX2B mutations, is reported in the histograms on the left: 25Ala, 29Ala, 33Ala (above) and c.930insG, c.614–618delC (below). Adapted by permission from Reference 18.

Moreover, based on electrophoretic mobility shift assays, it has been observed that expansions containing 29 alanines and more do affect PHOX2B DNA binding, probably because aggregated PHOX2B mutant proteins are not available for DNA binding, an observation confirmed in vitro by showing that PA expanded PHOX2B proteins spontaneously form oligomers (6). Finally, the interaction between the wild-type PHOX2B protein with the misfolded 33 repeats mutant has suggested that PA mutations can also prevent the normal protein from its usual function because of abnormal aggregation with the mutant (6, 18).

In the attempt to assess the fate of cells expressing PA-expanded PHOX2B, in vitro experiments have demonstrated that activation of the heat-shock response by the drug geldanamycin, a naturally occurring antibiotic, is efficient both in preventing formation and in inducing the clearance of PHOX2B preformed PA aggregates and, ultimately, also in rescuing the PHOX2B ability to transactivate the DβH promoter. In addition, elimination of PHOX2B mutant proteins by the proteasome and autophagy, two cellular mechanisms already known to be involved in the clearance of proteins containing expanded polyglutamine and polyalanine tracts, has been demonstrated. Cellular apoptosis has been observed only in association with the largest PA expansions (19).

Non-PA mutant PHOX2B proteins tested so far have shown compromised transcriptional activation of the DβH and TLX2 promoters with more severe activity disruption correlated with length of the frameshifted C-terminal sequence (see Figure 6B for the effect on the DβH target) (6, 18, 95). Unexpectedly, PHOX2B frameshift mutations have shown a 10 to 30% increased activation of the PHOX2A regulatory region (18). Moreover, frameshifts and missense mutations have mainly shown a complete loss of DNA binding but, unlike the long PA expansions, are able to correctly localize in the nucleus ( Figure 7 ) (6, 18).

Aberrant C-terminal regions may cause PHOX2B protein dysfunction due to either lack of ability to establish correct protein–protein interactions with molecular partners or gaining of the ability to interact with wrong molecules, a very attractive hypothesis in light of the association of NPARM mutations with risk of neuroblastoma development (96). Consistently, a recent study has shown that non-PA mutant PHOX2B constructs retained the ability to suppress cellular proliferation without being able to promote differentiation (20), suggesting a mechanism that might promote the development of neural crest tumors.

In conclusion, these studies have indicated a marked difference between PARMs and frameshift NPARMs in terms of transactivation of target promoters, formation of aggregates, and subcellular localization. Future in vitro investigation will provide further clues on pathogenesis and possible therapeutic hints.

Before the 2003 discovery that PHOX2B is the disease-defining gene for CCHS, the clinical spectrum of severity in terms of hypoventilation and other aspects of the ANSD phenotype had long puzzled clinicians. Once the diagnosis of CCHS is considered, blood should be sent for the PHOX2B Screening Test (see Figure 8 ). In the event the screening test is negative, and the patient's phenotype supports the diagnosis of CCHS or LO–CCHS or the physician/family wants to completely rule out the diagnosis of CCHS, then the sequel PHOX2B Sequencing Test should be performed (available at Children's Memorial Hospital, Chicago, IL for additional options please refer to www.genetests.org). This two-step testing is most cost efficient (the mutation in 95% of CCHS cases will be identified with the inexpensive PHOX2B Screening Test and only a subset of the NPARMs will require the PHOX2B Sequencing Test to be identified). While awaiting results of the clinically available PHOX2B testing (high sensitivity and specificity) other causes of hypoventilation should be ruled out to expedite proper intervention and facilitate treatment strategies for home care. Primary lung disease, ventilatory muscle weakness, and cardiac disease should be ruled out with the following tests: chest x-ray and potentially chest CT, comprehensive neurological evaluation and potentiallymuscle biopsy, and echocardiogram, respectively. Causative gross anatomic brain/brainstem lesions should be ruled out with an MRI and/or CT scan of the brain and brainstem (97, 98). Likewise, inborn errors of metabolism should be considered, and a metabolic screen should be performed.

Figure 8. Polyacrylamide gel electrophoresis PHOX2B Screening Test. Shown are PHOX2B polyalanine repeat expansion mutations (PARMs) by polyalanine repeat size and non-PARMs (NPARMs) for the most common congenital central hypoventilation syndrome (CCHS)-causing PHOX2B genotypes identified with the PHOX2B Screening Test compared with the wild-type (lanes 3 and 15) product. Lanes 1–2 indicate results for NPARMs with large deletions: 722del35 and 722del38. Lanes 4–14 indicate analysis of several PARMs, including mosaic parents (genotype 20/26 and 20/27) as well as probands (genotypes 20/24–20/33). Lanes 7 and 9 are mosaic carriers of CCHS-causing PARMs. Lane 7 is a mosaic parent of the proband identified in lane 6. Lane 9 is a mosaic parent of the proband identified in lane 8. Note that the band intensity of the expanded allele is much lighter than that of the wild-type allele in lanes 7 and 9. Also note that although the overall intensity of the signal in lane 8 is low, both the expanded and wild-type bands are similar in intensity indicating a full carrier of the expanded 27 repeat allele. Reproduced by permission from Reference 41.

Independent of respiratory control abnormalities, children with CCHS have other evidence for diffuse autonomic dysregulation (27, 28). For those individuals with constipation symptoms, a barium enema or manometry and potentially full thickness rectal biopsy should be performed to diagnose HSCR (99). Serial chest and abdominal imaging is essential among children with the NPARMs and those children with the 20/29 to the 20/33 genotype for emergence of a neural crest tumor, specifically neuroblastoma (NPARMs) and ganglioneuroblastoma/ganglioneuroma (PARMs) (100). Because no children with genotypes 20/24 to 20/28 have been identified with tumors of neural crest origin, the value of serial imaging in these cases is unknown. Cardiac rhythm abnormalities, including decreased beat-to-beat heart rate variability, reduced respiratory sinus arrhythmia, and transient abrupt asystoles, have been described (9, 101, 102). Seventy-two–hour Holter monitoring performed annually may determine aberrant cardiac rhythms, sinus pauses that will necessitate bipolar cardiac pacemaker implantation (103), and the frequency of shorter pauses (i.e., less than 3 s) that may have physiologic and neurocognitive impact. Children with CCHS are at risk for progressive pulmonary hypertension and cor pulmonale as a result of recurrent hypoxemia due to inadequate ventilator settings or tracheostomy caliber, unrecognized hypoventilation during spontaneous breathing while awake, excessive exercise with resultant physiologic compromise, or suboptimal compliance with artificial ventilation. As a result, echocardiograms, hematocrits, and reticulocyte counts performed every 12 months will provide information regarding potential cor pulmonale and polycythemia, with testing performed more frequently if clinically indicated and warranted. CCHS patients frequently exhibit ophthalmologic abnormalities reflecting the role of PHOX2B on the cranial nerves controlling pupillary function (23, 29). Comprehensive ophthalmologic testing will determine the nature of the ophthalmologic involvement and allow for intervention strategies to avoid interference with learning. Anecdotal reports of poor heat tolerance and profuse sweating have been described (1) but not studied comprehensively. Very limited formal assessment of the ANS has been reported, and none have been analyzed by PHOX2B genotype. Comprehensive autonomic testing as clinically indicated to assess syncope and to assess autonomic nervous system function may include tilt testing, deep breathing, Valsalva maneuver, thermal stressors, pupillometry, and more, as new measures of autonomic testing are developed for infants and children.

Suboptimal school performance and/or decreased intellectual function have been observed in CCHS patients (23, 104–107). It is unclear whether this is due to hypoxemia from inadequate ventilatory support or a direct result of the primary neurologic problem associated with CCHS. As children with CCHS are more consistently identified in the newborn period, and as management for these complex and vulnerable children becomes more standardized, improved neurocognitive performance is anticipated with distinction between sequelae of hypoxemia (due to hypoventilation or asystoles) and innate disease specific to CCHS. Comprehensive neurocognitive testing performed annually in a controlled setting will assess the child's progress relative to intervention, management, and compliance and may identify areas for intervention. Children with CCHS have a good future with the oldest neonatally identified patients graduating from college, getting married, and maintaining employment. It behooves the family and medical personnel to provide optimal oxygenation and ventilation to assure maximization of neurocognitive potential. Aggressive educational intervention coupled with careful ventilatory and cardiovascular management is essential (23, 104–107).

In the ideal situation, care of individuals with CCHS would be provided through centers with extensive expertise in CCHS, working in close partnership with regional pediatric pulmonologists and pediatricians. That arrangement will further improve the consistency of management and ideally improve the level of outcome for individuals with CCHS. Although there may be resistance to referral to such centers, the ATS Statement authors note that many exceedingly capable pediatric pulmonologists are basically managing children with CCHS as they would other patients with tracheostomies and ventilators. They are seeing the patients at varying intervals ranging from 4 to 12 months, only checking physiologic measures for 5 minutes awake in the clinic, not taking an autonomic medicine history, not reviewing the technology in the home or access to emergency power and care, not reassessing tracheostomy tube size after the neonatal period, not changing from neonatal ventilator tubing to pediatric tubing, not educating parents in ventilator management and usefulness of noninvasive monitoring, not advocating that all parents be screened with the PHOX2B Screening Test to ascertain mosaicism, and more. Essentially they are extremely capable but busy practitioners who do not have the time or intensive experience to focus on the nuances that are essential to the successful management of children with CCHS and their families. The aim is to accept that caring for children with CCHS is a privilege that requires a depth of care not typical for other patients with more common pulmonary disorders. There is simply not enough time for the pediatric pulmonologist (who is juggling patients with cystic fibrosis, chronic lung disease, asthma, and more) to provide in-depth attention to the needs of a child with CCHS. Because of the nature of the PHOX2B mutations, and the range of phenotype based on these mutations, experience with even 15 to 30 patients does not begin to provide the scope of experience necessary for understanding the needs of the child with CCHS.

Alveolar hypoventilation is the hallmark of CCHS, and its most apparent and potentially debilitating phenotypic feature. Characteristically, the diminution of tidal volume with resultant effect on minute ventilation is most apparent in non–REM sleep in CCHS, but it is also abnormal during REM sleep and wakefulness, although usually to a milder degree (23, 108, 109). The spectrum of sleep-disordered breathing may range in severity from hypoventilation during non–REM sleep with adequate ventilation during wakefulness, to complete apnea during sleep and severe hypoventilation during wakefulness. The CCHS phenotype relative to ventilatory needs is PHOX2B genotype/mutation-dependent. Typically, children with the 20/27 to 20/33 genotype and the NPARMs will require 24 hours per day of mechanical ventilation. Children with the 20/24 and 20/25 genotypes, and a small subset of NPARMs, rarely require 24 hours per day of ventilation unless they have had suboptimal ventilatory management for prolonged periods in early childhood. The awake ventilatory needs of the children with the 20/26 phenotype will vary with the activity level. It remains unclear whether spontaneous breathing while awake will improve with puberty.

The incidence of CCHS in the general population is unknown, although the incidence will likely vary by ethnicity based on current statistics (∼90% of identified subjects are Caucasian in review of published reports and in personal communication from the laboratories of the members of the ad hoc Statement Committee). With recognition that individuals with the 20/24 and 20/25 genotypes have variable expressivity, and might not present to medical attention until they receive sedation or have severe pulmonary illness, it behooves the medical community to identify such patients before being faced with a life-threatening situation. What is apparent is that the incidence of CCHS is no longer as rare as anticipated at the time of the first ATS Statement on CCHS in 1999 (1). Determination of a true incidence will require a large population-based study across all ethnic groups.

The results from reports of awake and asleep ventilatory and arousal responses (108, 110–112), mental concentration (34), respiratory sensations (35, 36), physiologic response to exercise and leg motion (32–35, 113, 114), and focal abnormalities on functional MRI (115–118) must be interpreted with caution as they likely reflect bias due to small sample size, reporting in the pre-PHOX2B era, data presentation inclusive of subjects with CCHS and with other causes of hypoventilation (108), or without documentation of specific PHOX2B confirmation of CCHS in all subjects, and near-exclusive inclusion of children who were able to sustain adequate ventilation during wakefulness at rest (so likely nearly all subjects had the 20/25 genotype based on the described phenotype and protocols). Until the studies are repeated in large cohorts including a broad array of children and adults with genotypes 20/24 to 20/33 as well as the NPARMs, these results may represent the physiology of only the mildest patients with CCHS.

Biannual then annual in-hospital comprehensive physiologic studies during awake and asleep states to assess ventilatory needs during varying levels of activity and concentration, in all stages of sleep, with spontaneous breathing and with artificial ventilation, and ventilatory responsivity to endogenous and exogenous physiologic challenges awake and asleep, will ascertain each child's needs for optimal clinical management. These studies, performed over the course of a several-day hospitalization in a center with extensive CCHS experience, will allow for a clear understanding of needs when breathing spontaneously as well as with artificial ventilation by any of the means described in the sections that follow. These physiologic studies should include constant supervision by highly trained personnel and continuous audiovisual surveillance with continuous recording (at a minimum) of respiratory inductance plethysmography (chest, abdomen, sum), ECG, hemoglobin saturation, pulse waveform, end tidal carbon dioxide, sleep state staging, blood pressure, and temperature. Other recommended testing for individuals with PHOX2B mutation-confirmed CCHS is provided in Table 1.

TABLE 1. RECOMMENDED TESTING TO CHARACTERIZE CCHS PHENOTYPE

Definition of abbreviations: PARM = polyalanine repeat expansion mutation NPARM = nonpolyalanine repeat expansion mutation (missense, nonsense, frameshift).

* Infants under the age of three years should undergo comprehensive evaluations every 6 months.


In insects, does the Alanine repeat occur on the homeodomain sequence of the abdomen or does it occur on a different sequence? - Biology

Comment: reported as procephalic ectoderm anlage

Comment: reported as procephalic ectoderm anlage

Comment: reported as procephalic ectoderm anlage

Comment: reported as procephalic ectoderm anlage

Comment: expressed in female fat tissue surrounding the spermothecae

Expression was examined at four phases of embryonic stage 5. prd is initially expressed in a nonperiodic gap-like pattern in the anterior, which begins to resolve into stripes in phase 2 however, the full 7-stripe pattern arises only during phase 3. The stripes emerge fully refined, with sharp, evenly spaced stripes.

prd expression is enriched in adult males.

prd transcription is repressed by ectopic eve protein in eve hs.PS embryos. Timing suggests that eve is a direct regulator of prd.

prd transcripts rise to a sharp peak in 2-4 hr embryos, rapidly decline and are undetectable after 12 hrs. They are absent in oocytes. prd is initially expressed with double segment periodicity (7 stripes) but switches at cellular blastoderm to a pattern of single segment periodicity (14 stripes).

A 3.6kb prd transcript is observed in addition to the 2.5kb transcript in 0-4hr embryo RNA from prd 4 heterozygous mothers.

prd protein is expressed in male accessory glands. It is initially expressed at high levels in all accessory gland cells, but protein levels decline with increasing age of virgin males. prd protein levels decline more rapidly in the main cells than in the secondary cells. In 10 day males, prd protein is detected only in scattered secondary cells in the distal region of the glands. Mating increases prd protein level

s in both main and secondary cells throughout the glands.

prd protein is first detected in the anterior region of very early embryos and is subsequently restricted to a very narrow anterior stripe. During cellularization, bell-shaped stripes (stripes 3-7) are observed. The order of stripe appearance is 4 and 7, followed by 3 and 6, and finally stripe 5 emerges. At this point the anterior stripe splits into two. By mid-cellularization prd protein has reached equal levels in all stripes. At the same time, the preferential accumulation of prd protein at the posterior margins generates a gradient within each stripe. During the second half of cellularization, expression also occurs in a patch of cells at the anterior dorsal end of the embryo and in an eighth stripe. By the end of gastrulation, the number of stripes has doubled to 14 as a result of reduction of protein in the middle of the stripes accompanied by an increase of protein in the anterior region of the stripes. At the end of the germ band extended stage, the expression in the stripes has disappeared. prd protein is expressed in the head region, most strongly in the maxillary lobe, but also in the labial and mandibular lobes and in the clypeolabrum. prd protein is also detected in the developing CNS in 2-3 specific neurons per hemisegment.


Classification of Hox proteins

Proteins with a high degree of sequence similarity are also generally assumed to exhibit a high degree of functional similarity, i.e. Hox proteins with identical homeodomains are assumed to have identical DNA-binding properties (unless additional sequences are known to influence DNA-binding). To identify the set of proteins between two different species that are most likely to be most similar in function, classification schemes are used. For Hox proteins, three different classification schemes exist: phylogenetic inference based, synteny-based, and sequence similarity-based. [20] The three classification schemes provide conflicting information for Hox proteins expressed in the middle of the body axis (Hox6-8 and Antp, Ubx and abd-A). A combined approach used phylogenetic inference-based information of the different species and plotted the protein sequence types onto the phylogenetic tree of the species. The approach identified the proteins that best represent ancestral forms (Hox7 and Antp) and the proteins that represent new, derived versions (or were lost in an ancestor and are now missing in numerous species). [21]


In insects, does the Alanine repeat occur on the homeodomain sequence of the abdomen or does it occur on a different sequence? - Biology

Comment: reported as procephalic ectoderm anlage

Comment: reported as procephalic ectoderm anlage

Comment: reported as procephalic ectoderm anlage

Comment: reported as procephalic ectoderm anlage

Comment: expressed in female fat tissue surrounding the spermothecae

Expression was examined at four phases of embryonic stage 5. prd is initially expressed in a nonperiodic gap-like pattern in the anterior, which begins to resolve into stripes in phase 2 however, the full 7-stripe pattern arises only during phase 3. The stripes emerge fully refined, with sharp, evenly spaced stripes.

prd expression is enriched in adult males.

prd transcription is repressed by ectopic eve protein in eve hs.PS embryos. Timing suggests that eve is a direct regulator of prd.

prd transcripts rise to a sharp peak in 2-4 hr embryos, rapidly decline and are undetectable after 12 hrs. They are absent in oocytes. prd is initially expressed with double segment periodicity (7 stripes) but switches at cellular blastoderm to a pattern of single segment periodicity (14 stripes).

A 3.6kb prd transcript is observed in addition to the 2.5kb transcript in 0-4hr embryo RNA from prd 4 heterozygous mothers.

prd protein is expressed in male accessory glands. It is initially expressed at high levels in all accessory gland cells, but protein levels decline with increasing age of virgin males. prd protein levels decline more rapidly in the main cells than in the secondary cells. In 10 day males, prd protein is detected only in scattered secondary cells in the distal region of the glands. Mating increases prd protein level

s in both main and secondary cells throughout the glands.

prd protein is first detected in the anterior region of very early embryos and is subsequently restricted to a very narrow anterior stripe. During cellularization, bell-shaped stripes (stripes 3-7) are observed. The order of stripe appearance is 4 and 7, followed by 3 and 6, and finally stripe 5 emerges. At this point the anterior stripe splits into two. By mid-cellularization prd protein has reached equal levels in all stripes. At the same time, the preferential accumulation of prd protein at the posterior margins generates a gradient within each stripe. During the second half of cellularization, expression also occurs in a patch of cells at the anterior dorsal end of the embryo and in an eighth stripe. By the end of gastrulation, the number of stripes has doubled to 14 as a result of reduction of protein in the middle of the stripes accompanied by an increase of protein in the anterior region of the stripes. At the end of the germ band extended stage, the expression in the stripes has disappeared. prd protein is expressed in the head region, most strongly in the maxillary lobe, but also in the labial and mandibular lobes and in the clypeolabrum. prd protein is also detected in the developing CNS in 2-3 specific neurons per hemisegment.


Methods

Insects and materials

Anopheles gambiae (strain PEST) mosquitoes were raised and maintained in an environmental chamber at 26°C, 85% relative humidity, with a 16-hour light, eight-hour dark cycle including a one-hour dusk/dawn period [18]. Larvae were fed daily a 2:1 mixture of fish pellets: brewer’s yeast, that had been finely ground [19]. DL-octopamine, tyramine, dopamine, naphazoline, clonidine, serotonin, chlorpromazine, cyproheptadine, promethazine, all hydrochloride salts, and tolazoline a benzylimidazoline salt, were obtained from Sigma-Aldrich. Metoclopramide hydrochloride was obtained from MP Biomedical. Compounds identified in the virtual screen were purchased from Princeton BioMedical, ChemDiv, Chembridge and Enamine and tested in vitro against AgOAR45B expressed in the GloResponse™CRE-luc2P HEK293 reporter cell line and in larval bioassays.

Expression analysis of AgOAR45A and AgOAR45B

Total RNA was isolated from five different An. gambiae immature stages (L1-P), adult females and males, adult female heads only, and adult female abdomen/thorax using the RNeasy Mini Kit (Qiagen). The DNase (Fermentas)-treated RNA was used to generate cDNA using Superscript III (Invitrogen) and oligo (dT12–20), according the manufacturer’s recommendations. Quantitative PCR (qPCR) was performed using SYBRGreen (ABI), an ABI 7900 RT-PCR system and 200 ng of cDNA per sample, a final concentration of 0.15M of each primer, and an annealing temperature of 60°C. Primer sets used for expression analysis were: Ag10592 forward- CACCATCGAACACAAAGTTGACACTT Ag10592 reverse- CGAACGTAACGTCACGGCCA Ag45A&B forward- GGGTACGTCGTCTACTCAGCCCTC Ag45A reverse- TGTATCCGCAGCGTTAGCCGATTG Ag45B reverse- CGAGATTGTTCTTGCCACCTTTGGTG. The 40S Ribosomal protein subunit 7 (AGAP01592) was used as an internal control. Reactions for each gene and for the control used were carried out in triplicate. Relative expression levels of each gene was determined by the ΔΔCT method, where relative expression is expressed as a fold difference relative to whole females and expressed as 2 - ΔΔCT . The following formula was used: ΔΔCT = ΔCT(stage or condition) − ΔCT(Females) and ΔCT = CT (gene of interest) − CT (40S RNA).

Heterologous expression of AgOAR45B octopamine receptor

Total RNA was isolated from heads of three-day old adult females using RNeasy Mini Kit (Qiagen). cDNA was synthesized using SuperScript III (Invitrogen), and used as a template for PCR amplification of the AgOAR45A and AgOAR45B genes. Insertion of the coding sequences into the Sgf I and Pme I sites of the pF9a CMV hRluc-neo vector (Promega) was performed by digestion of a fragment amplified by the following primers: Ag45forward- TAAAGCGATCGCCATGAACGAGTCGGAGTGTGCC Ag45Areverse- TTGTGTTTAAACTCTCGAGTCGGACAGGTCGC Ag45Breverse- CGCGGTTTAAACTCTGAACACACCACCGACGA. Primers were constructed based on the annotated sequence of the AGAP000045 gene (VectorBase, Protein ID: AGAP000045-PA and -PB [20].

GloResponse™CRE-luc2P HEK293 reporter cell line (Promega) were maintained as adherent culture at 37°C, 5% CO2 in DMEM (Invitrogen) supplemented with 10% fetal calf serum (Atlanta), and 50 mg/ml hygromycin B. Transfection of cells was carried out using the Amaxa Nucleofector kit per the manufacturer’s instructions. Control transfections were performed using a pF9A plasmid with the barnase (Bacterial Ribonuclease) gene removed as suggested by the manufacturer (Promega). Stable lines were created by applying 400 mg/ml G418 for three weeks. Stable clones of AgOAR45B expressing HEK293 reporter cells were created through two rounds of limiting dilution cloning.

CAMP assay

Intracellular cAMP increase was monitored through a CRE-luc reporter construct in HEK293 cells (Promega). Stable cell lines were plated at a density of 4×10 4 cells per well in white 96-well assay plates (Corning, Cat. #3917). Cells were immediately treated with the various compounds, and returned to the incubator for four hours before being assayed. cAMP was quantified through luciferase production using the Dual-Glo luciferase assay kit (Promega) following the manufacturer’s instructions. Luciferase units were normalized to the cell number using the internal Rluc construct in the pF9A expression plasmid.

Intracellular Ca ++ assay

Stable cell lines expressing AgOAR45A or AgOAR45B were plated in black well, clear bottom, 96-well assay plates (Greiner Bio-One, Cat. #655090) at 4×10 4 cells per well. Cells were allowed to grow overnight before being assayed. Cells were preloaded with Fluo-4 NW (Molecular Devices), and carried out per the manufacturer’s instructions. Fluorescence was monitored before and after addition of compounds using the Flexstation3 (Molecular Devices), at two-second intervals for 120 seconds.

Site-directed mutagenesis of AgOAR45B

The AgOAR45B gene was subcloned into a TA vector (Invitrogen), according to the manufacturer’s instructions. Site-directed mutagenesis of AgOAR45B was carried out using Quick Change Lightning Kit (Agilent). Primers (IDT) were designed to introduce single amino acid changes (see Additional file 1). Double mutants were created through two rounds of mutagenesis. Mutations were confirmed through automated sequencing (ND Genomics Core Facility). Mutant genes were then excised and moved to the pF9A expression vector as described above.

Membrane preparation and radioligand binding assay

Cells in T75 flasks were washed with 10 ml of PBS, removed by scraping, centrifuged at 500 g, and then resuspended in 5 ml of lysis buffer, 50 mM Tris-Cl pH 7.4. Cells were incubated on ice for 10 min and homogenized with a Dounce homogenizer with 30 strokes. The homogenate and 5 ml wash with lysis buffer were then centrifuged at 23,000 g for 30 min at 4°C to pellet crude membranes. The resulting pellet was re-suspended in 50 mM Tris-Cl pH 7.4, 5 mM MgCl2, 0.5 mM EDTA pH 7.4 and quantified by micro BCA assay (Thermo).

Radioligand binding assays were carried out using 3 H-Yohimbine to determine binding affinity of the octopamine receptor with different compounds. Isolated membranes, 30 μg per well, were incubated in the presence of 16 nM 3 H-Yohimbine (American Radiolabeled Chemicals) in binding buffer, and various concentrations of compounds. The final reaction volume was 125 ml per well in a 96-well assay plate. One-hundred mM clonidine was used to determine non-specific binding in each experiment. After a two-hour incubation, membranes were collected and washed on filter mats, pretreated with 0.3% polyethylenimine (Sigma) using a Brandel 96-well harvester and counted in a Perkin Elmer Microbeta counter.

Larval bioassay

Dose response curves were made to determine the LD50 for some compounds against three-day old Aedes aegypti larvae. Ten larvae were put in each well of a 12-well plate, with each well containing a different concentration of compound in water. Three replicate wells were made for each experiment and the curves performed three separate times. Control wells were included in each experiment, which contained only DMSO in water. Plates were incubated 24 hours in standard insectary conditions. Mortality was determined after 24 hours.

Project workflow

A computational-experimental workflow (see Additional file 2), similar to the computational approach described by Yarnitzky et al.[21] was utilized. The method was as follows: 1) homology models, both inactive (antagonist-based) and active (agonist-based) conformations, were created using the fragment-based method I-TASSER (Iterative Threading Assembly Refinement) [22–24] as well as surrounded by lipids, water, and ions using visual molecular dynamics (VMD) [25] 2) a preliminary virtual screen with one of two compound test sets (composed of either GPCR agonists or antagonists whose activity on AgOAR45B was experimentally determined using an AgOAR45B reporter assay) was performed and a set of top-ranked positions of the most were chosen and used in MD simulations with the proper active (agonist-binding) and inactive (antagonist-binding) protein conformations 3) MD simulations were performed until stable ligand positions were obtained, then an additional virtual screen with the test set was performed on each resulting protein conformation and the results were analysed manually. Step 3 was repeated until the ligand positions were stabilized in the protein even after ten additional nanoseconds (ns) of simulation 4) final conformations were then used to build grids which were subjected to virtual screening using the ZINC library and, 5) the resulting compounds were analysed and compounds with differing structural characteristics were purchased and experimentally tested in vitro.

Homology modelling

Both initial inactive (antagonist-based) and active (agonist-based) conformations of AgOAR45B were generated using the I-TASSER online server [26]. For the inactive conformation, a structure was obtained that was built based on many GPCR antagonist-bound and inverse agonist-bound conformations, primarily a crystal structure of β2-adrenergic receptor with partial inverse agonist carazolol bound [PDB:2RH1]. For the active conformation, a structure was obtained that was built based on the crystal structure of the β2-adrenergic receptor-Gαs protein complex with high affinity agonist BI-167107 bound [PDB:3SN6]. Both conformations were obtained using standard I-TASSER settings. In the case of the active conformation of AgOAR45B, 3SN6 (chain R) was used as a template to generate the model. Root-mean-square deviations (RMSDs), as a measure for protein stability, were calculated using the VMD 1.9 RMSD Trajectory Tool [25]. AgOAR45B has a 24% sequence identity (35% similarity) to the β2-adrenergic receptor (see Additional file 3). The active site residues were determined from analyzing the inactive and active β2-adrenergic receptor’s crystal structures and observing residues that interacted or had the potential to interact with the crystal structures’ bound ligands [27, 28]. The active site sequence identity is much higher at 67% (73% similarity).

Molecular dynamics

Using VMD 1.9, each virtual representation of the protein was first embedded in a large 1-palmitoyl-2-oleoylphosphatidylcholine (POPC) lipid bilayer, removing any lipid that overlapped with the protein. The virtual protein-membrane system was then solvated with TIP3P water molecules, and neutralized by virtually adding KCl up to 400 mM. Initially, CHARMM 27 parameters [29] were assigned to all molecules using VMD 1.9 to enable the addition of the lipid bilayer, water and ions. However, once each virtual system (both active and inactive conformation) was prepared and its respective ligand (octopamine for the active conformation and promethazine for the inactive conformation) was ready to be added, AMBER gaff and ff03.r1 parameters were assigned to the ligand and the rest of the molecules, respectively, using Amber 11 tleap [30–32]. The AMBER force field was chosen as it allows the generation of parameters for the ligand using the antechamber module [33]. A disulphide bridge was added between the residues of Cys93 and Cys194, as this bridge also existed in the template PDBs.

Each complete virtual system consisted of the respective conformation of AgOAR45B embedded to a large POPC bilayer with 168 lipid molecules. In each virtual system, all residues were at the normal protonation state for physiological pH. In addition, the antagonist- bound system contained 171 potassium ions, 204 chloride ions, and 22,525 water molecules for a total of 99,706 atoms (measured 95×96×128 Å), while the agonist-bound system contained 172 potassium ions, 205 chloride ions, and 22,654 water molecules for a total of 100,077 atoms (measured 101×92×127 Å). Before MD simulations the systems were equilibrated using 120 CPU cores as follows: 1) MD of lipid tails for 500 picoseconds (ps) [time step = 2 femtoseconds (fs)] with protein, ligand, lipid head groups, water, and ions kept fixed 2) equilibration for lipids, water and ions for 500 ps (time step = 2fs) with harmonic constraints on the protein and ligand 3) equilibration of the entire system for 500 ps (time step = 2fs) with no molecular constraints. After equilibration, 20–30 ns of MD simulation were performed using 504 CPU cores in two to three 10 ns increments, with time step = 2fs and trajectory data being collected every 200 ps. The equilibration and simulation steps were run using NAMD 2.8 [34] on the high-performance computing cluster Kraken [35].

Virtual screen preparation

All virtual screening jobs were run on their respective ‘protein only’ homology models (i.e., containing no lipids, waters or ions). The proteins were prepared by first running the Protein Preparation Wizard workflow [36, 37], and then grids were generated using Glide’s Receptor Grid Generation application, each found in Schrodinger Suite 2011’s Maestro [38]. To obtain initial ligand poses used in grid generation, each of the conformations were first overlapped in PyMOL [39] with the top templates used by I-TASSER for their creation (2RH1 for the inactive conformation and 3SN6 for the active conformation) and the positions of the ligands found in each of the templates were first saved as a PDB file, then added using Schrodinger Suite 2011’s Maestro to their respective AgOAR45B conformation (2RH1 partial inverse agonist CAU was used for the inactive conformation and 3SN6 agonist 30G was used for the active conformation) to denote the active site of the protein. The ligand added to each of the homology models was used as the centroid of the grid, determining the area in which the libraries of compounds should be docked. No constraints were used.

Virtual screening with known GPCR ligands

The inactive and active conformations of AgOAR45B with the added ligands were used to build the grids, which were then run in a virtual screen using Schrodinger’s Glide software [40–43] against known GPCR antagonists and agonists, respectively. The original compound structures used in the two test sets were downloaded from the NIH’s PubChem website. One of the test sets contains known GPCR agonists: synephrine, cinnamic acid, clonidine, demethylchlordimeform, dopamine, eugenol, histamine, naphazoline, norepinephrine, octopamine, phentolamine, serotonin, tolazoline, transanethole, and tyramine. The other containing known GPCR antagonists: rauwolscine, demethylchlodimeform, 8-hydroxymianserin, amitriptyline, antazoline, chlorpromazine, cyproheptadine, desipramine, desmethylmianserin, dihydroergotamine, gramine, imipramine, maroxepin, metoclopramide, mianserin, phentolamine, prazosin, promethazine, propranolol, triprolidine, and yohimbine. Each of the test sets was then prepared using Schrodinger Maestro’s LigPrep [44] to generate different potential protonation states at pH range 7–8, tautomers and ring conformations. After each of the test sets was docked to its respective AgOAR45B conformation, top pose positions for the antagonist promethazine and the agonist octopamine were saved and later added to the inactive (antagonist-bound) and active (agonist-bound) conformations, respectively. These protein conformations with docked ligands were then used as the starting position for the MD simulations.

Virtual screening with ZINC library

The library used in the docking contained drug-like compounds from the ZINC online database [45]. Compounds were prepared using Schrodinger Maestro’s LigPrep to generate different potential protonation states at pH range 5–9, tautomers and ring conformations. The final library contained approximately 12 million compounds. It was split into five sublibraries ( ∼ 2.4 million compounds per sublibrary) for the first run of high-throughput virtual screening.

The library was screened against five grids of AgOAR45B (two from the antagonist-bound conformations and three from the agonist-bound conformations), built from the receptor positions after 20 ns of MD simulation each. Virtual screening using the ZINC library was performed using Schrodinger’s Glide software in three steps: 1) each of the five sublibraries was filtered using high-throughput virtual screening and the top 30,000 compounds in each sublibrary were saved and recombined making a library containing 150,000 compounds 2) this new library was then filtered further using standard precision virtual screening and the top 15,000 compounds were saved and, 3) this new library was then filtered one more time using extra precision virtual screening and the top 1,500 compounds were saved and analysed, considering the score, the relevant interactions in the active site pocket and diversity in the sampling for further testing.


DISCONNECTED

Lisa Kay Robertson

A dissertation submitted to the Graduate Faculty of North Carolina State University

In partial fulfillment of the Requirements for the degree of

Doctor of Philosophy. Genetics

Dr. James W. Mahaffey, Chair of Advisory Committee,

Dr. Michael D. Purugganan, Committee Member,

Dr. Jonathan M. Horowitz, Committee Member, Molecular Biomedical Sciences

This work is dedicated to my amazing family, whose unconditional love and support make it possible for me to conquer all challenges in my path.

To my Dad, who is my constant inspiration and who so shaped me through his life of excellence, his unwavering character, and his never-ending desire to learn.

I recognize and extend my sincere thanks to the Department of Genetics for the amazing opportunity they have provided. I am very grateful for the advice and guidance from my advisor, Jim Mahaffey, and the members of my advisory committee—Amy Bejsovec, Michael Purugganan, and Jon Horowitz.

I must also thank all members of the Mahaffey Lab, past and present, for their friendship and assistance. In particular, I thank Kate Lingerfeld, Dana Bowling, Jamie Mahaffey, and Mark Shepherd for their efforts and contributions to my research Muk Patel for his superior molecular skills and steel trap memory Barbara Imiolczyk for all of her advice and our wonderfully efficient lab manager Jennifer Hutchinson, for all she does to keep the lab running.

I must also express my appreciation to members of the Brown and Denell labs at Kansas State University for their generous sharing of DNA, Tribolium protocols, and beetle wrangling tips. I also thank Bijan Dey in the Campos lab at McMaster University for sharing authorship, fly lines, and information regarding disconnected.

CHAPTER ONE: General Introduction. 1

CHAPTER TWO: Insect HOM-C Genes and Development —Lessons from Drosophila and Beyond . 5

Introduction—Drosophila Hom-C/Hox Genes . 7

The Embryonic Expression and Mutant Phenotypes of Deformed and Sex combs reduced in Drosophila and Tribolium . 16

Initiation of Deformed and Sex combs reduced Expression in Drosophila . 22

Maintenance of Dfd and Scr Expression in Drosophila . 23

Post-translational Regulation of SCR. 26

Regulation between the HOM-C Proteins—Auto and Cross-regulation of DFD and SCR . 26

The Hom-C Protein Homeodomain. 28

Hom-C Protein Function . 30

Functional Specificity of the Hom-C Homeodomain . 32

Cooperative Interactions Between Hom-C Proteins and Other Factors . 35

Extradenticle is a Hom-C Protein Co-factor . 38

The YPWM (Hexapeptide) Motif and Exd/Hom-C Heterodimer Formation. 41

Homothorax Forms a Trimer with Exd and Hom-C Proteins. 42

As a Co-factor, Extradenticle is Insufficient to Fully Explain Hom-C Protein Function . 42

Other Potential Hom-C Protein Co-factors-- cap’n’collar, lines, and apontic. 44

Potential C2H2 Zinc Finger Protein Co-Factors. 45

The Focus of This Research. 48

CHAPTER THREE: Expression Of The Drosophila Gene Disconnected Using the UAS/GAL4 System

CHAPTER FOUR: An interactive network of zinc finger proteins contributes to regionalization of the Drosophila embryo and establishes the domains of Hom-C protein function. 81

Materials and Methods . 86

Both Disco and Dfd are required to specify maxillary identity. 87

Tsh represses Disco during normal trunk segment development . 90

Ectopic Disco alters trunk development. 95

Disco and Scr can activate Scr target genes . 100

CHAPTER FIVE: Expression of the Drosophila gap gene giant establishes the gnathal trunk boundary via the activation and positioning of the C2H2 zinc finger tagmosis factors disconnected and teashirt. 114

Materials and Methods . 119

disco mRNA distribution is altered in early acting patterning gene mutants. 120

tsh mRNA distribution is altered in gap mutants . 125

The alterations in disco and tsh expression in gt mutants suggest a homeotic transformation of the labial segment . 129

Homeotic Gene Expression is Altered in gt Mutant Embryos . 136

tsh is required for the gnathal to trunk transformation in gt mutants . 139

Antp is not required for the ectopic epidermal expression of tsh in the gnathal segments . 140

The anterior gt domain functions in segment identity, in addition to segmentation . 142

The function of the anterior gt domain in Drosophila is similar to Tribolium. 145

gt expression establishes the gnathal/trunk boundary via the regulation of tsh and disco. 146

CHAPTER SIX: Functional characterization of the disconnected homologue in the Coleopteran beetle, Tribolium castaneum. 159

Materials and Methods . 163

Isolation, structure and evolutionary conservation of Tc’disco . 165

Tc’disco is expressed in the appendages during embryonic development. 167

Tc’disco is required for proper embryonic appendage patterning and elongation . 178

Tc’disco is required for proper adult appendage patterning and elongation . 181

Tc’Disco is a member of an evolutionarily conserved C2H2 zinc finger family . 182

Embryonic Tc’disco expression is similar to the embryonic expression of Dm’disco. 185

Tc’disco functions in Tribolium appendage development. 188

Possible Roles for Tc’disco in Tribolium appendage development . 189

CHAPTER SEVEN: Summary And Prospects. 199

Summary And Looking Forward . 200

Conclusions and Suggestions for Future Work. 201

Table 1: The effects of examined genes on disco mRNA accumulation. 121

Figure 1: The molecular organization of the Drosophila Homeotic Complex

Figure 2: Phenotypic consequences of Insect Hom-C Mutations . 10

Figure 3: Regional Hom-C gene expression in Drosophila embryos . 13

Figure 4: Segments, Parasegments, and Compartments. 15

Figure 5: Structures of the Drosophila larval head . 18

Figure 6: The Hom-C protein homeodomain . 31

CHAPTER THREE Figure 1: Embryonic in situ mRNA localization of endogenous disco and UAS-discoM expression . 73

Figure 2: Phase contrast microscopy images of wild type and UAS-discoM, armadillo-GAL4 cuticle preparations. 75

Figure 3: Environmental scanning electron micrographs (ESEM) of compound eyes comparing wild type and UAS-discoC expressing individuals . 77

CHAPTER FOUR Figure 1: Disco and Dfd are required for maxillary identity . 88

Figure 2: Trunk to gnathal transformation is more complete in tsh mutant embryos. 91

Figure 3: The role of Tsh in disco mRNA distribution . 93

Figure 4: Hom-C proteins accumulate in the proper register in embryos with ectopic Dfd and Disco but lacking Tsh . 96

Figure 5: Dorsal closure is blocked by prd » disco expression . 98

Figure 6: Ectopic Disco alters the trunk segments. 101

Figure 7: Activation of the Scr target gene, PB, by ectopically expressed disco and Scr. 103

Figure 8: An interactive hierarchy of zinc finger transcription factors establishes trunk and gnathal/head segment types . 109

CHAPTER FIVE Figure 1: The spatial expression of disco is affected by mutations in hb, Kr, and gt. 122

Figure 2: tsh mRNA accumulation in hb, Kr, and gt mutant embryos. 126

Figure 3: Cuticle analysis of hemizygous gt mutant embryos reveals homeosis of the labial segment . 131

Figure 5: Alterations is Hom-C/Hox Gene Expression in gtQ292 mutant embryos

indicate a homeotic change in labial identity. 138

Figure 6: tsh is required for the homeosis of the labial segment in gtQ292 mutants. 142

Figure 7: The gap gene giant has both homeotic functions and segmentation functions. 151

CHAPTER SIX Figure 1: The predicted protein sequence for Tc’Disco. 166

Figure 2: mRNA accumulation of Tc’disco in the Tribolium embryos. 168

Figure 3: Cuticle preparation of Tc’disco RNAi larval legs. 172

Figure 4: Cuticle preparation of Tc’disco RNAi larval head appendages. 175

Figure 5: Tc’disco is required for proper patterning and elongation of the adult appendages. 179

Figure 6: A phylogenetic analysis of zinc finger regions from the Disco family of proteins . 184

Figure 7: Comparison of disco transcript accumulation, at the germband extended stage, between Drosophila and Tribolium. 186

Figure 8: A possible model for Tc’disco regulation and function in larval appendage development. 192

CHAPTER SEVEN Figure 1: Possible Mechanistic Models for Hox/Zinc Finger control of target gene transcription. 204

CHAPTER ONE

Although mutations in the Drosophila melanogaster (the fruit fly) Homeotic Complex (Hom-C) were previously isolated, it was not until a 1978 Nature report by Edward B. Lewis that these genes were recognized as being part of a complex of genes required for proper anterior-posterior patterning of the Drosophila embryo (Lewis 1978). Further work by the Kaufman and McGinnis labs expanded the number of identified Hom-C genes to eight (Kaufman et al. 1980) and identified a highly conserved sequence called the homeobox. This sequence encodes a DNA binding motif, called the homeodomain (McGinnis et al. 1984). Since, there has been extensive research into the conservation and function of these transcription factors in many different animal species (Reviews in McGinnis and Krumlauf 1992 Gellon and McGinnis 1998 Veraksa et al. 2000). These genes (generally called the Homeobox (Hox) genes) are found to play a crucial role in the embryonic development and patterning in virtually all animals, and likely underlie both developmental and evolutionary diversity in the animal kingdom (Castelli-Gair 1998 Hughes and Kaufman 2002 Prince 2002). The Hox genes encoding these proteins are conserved in a clustered genomic arrangement that is usually co-linear with their anterior/posterior domains of expression. Also called “selector” genes (García-Bellido 1977), the Hox genes encode transcription factors presumed to regulate specific genetic cascades to confer precise and unique regional identities along the anterior/posterior body axis. Because different Hox proteins are able to bind very similar, and sometimes identical, target DNA sequences, it is likely that co-factors play a role in guiding Hox protein functional specificity during animal development. However, few such co-factors have been identified.

the function of disco during embryonic development, and how do the encoded Disco proteins function with the Hox proteins to direct patterning? Second, what upstream factors regulate the spatial and temporal expression of disco and disco-r during embryonic development? Finally, given that disco homologues are found in diverse animal phyla, as are the Hox genes, is the Hox co-factor function also conserved? These questions are examined using the model organisms Drosophila melanogaster and Tribolium castaneum (the red flour beetle).

The next chapter of this work gives a comprehensive overview of Hom-C/Hox protein function in Drosophila and other insects, and describes the questions that remain regarding Hox protein function. Following is the presentation of the results from ectopic expression and genetic analysis utilized to explore disco function in Drosophila. disco expression is shown to define domains in which Dfd and Scr can function. disco is further shown to have a regional embryonic patterning role as part of an interactive network of C2H2 zinc finger proteins that includes the trunk patterning gene and trunk Hox co-factor, teashirt (tsh). The results of this work leads to the proposal of a model for Hox protein function in Drosophila in which different combinations of regionally expressed C2H2 zinc finger proteins and Hox proteins confer specific segment identities along the embryonic body axis. Next, the results of a screen for upstream regulators of disco and tsh are presented, and the gap gene giant (gt) is shown to be key in setting the boundaries of expression of these two genes to define the Drosophila embryonic gnathal/trunk boundary. The regulation of disco and tsh by Gt underlies a previously unrecognized segment identity role for the Gt protein in its anterior expression domain. Finally, the identification and characterization of a disco homologue from Tribolium is presented. The embryonic expression pattern of the Tribolium disco is examined and compared to Drosophila disco expression. The embryonic and adult developmental functions of the Tribolium disco gene are investigated utilizing parental and larval RNA interference, and the results reveal a significant role for the Tribolium gene in appendage patterning and elongation. This work is concluded with a summary of the work, conjecture on the evolutionary conservation of disco function, and prospects for future work.

Castelli-Gair, J. (1998). Implications of the spatial and temporal regulation of Hox genes on development and evolution. Int J Dev Biol. 42: 437-444.

García-Bellido, A. (1977). Homoeotic and atavic mutations in insects. Amer. Zool. 17: 613-629.

Gellon, G. and McGinnis, W. (1998). Shaping animal body plans in development and evolution by modulation of Hox expression patterns. Bioessays 20(2): 116-125.

Heilig, J. S., Freeman, M., Laverty, T., Lee, K. J., Campos, A. R., et al. (1991). Isolation and characterization of the disconnected gene of Drosophila melanogaster. Embo J 10(4): 809-815.

Hughes, C. L. and Kaufman, T. C. (2002). Hox genes and the evolution of the arthropod body plan. Evolution &amp Development 4(6): 459-499.

Kaufman, T. C., Lewis, R. and Wakimoto, B. (1980). Cytogenetic Analysis of Chromosome 3 in Drosophila melanogaster: the Homoeotic gene complex in Polytene Chromosome Interval 84 A-B. Genetics 94: 115-133.

Lewis, E. B. (1978). A gene complex controlling segmentation in Drosophila. Nature 276(5688): 565-570.

Mahaffey, J. W., Griswold, C. M. and Cao, Q. M. (2001). The Drosophila genes disconnected and disco-related are redundant with respect to larval head development and accumulation of mRNAs from deformed target genes. Genetics 157(1): 225-236.

McGinnis, W. and Krumlauf, R. (1992). Homeobox genes and axial patterning. Cell 68(2): 283-302.

McGinnis, W., Levine, M. S., Hafen, E., Kuroiwa, A. and Gehring, W. J. (1984). A conserved DNA sequence in homoeotic genes of the Drosophila Antennapedia and bithorax complexes. Nature 308(5958): 428-433.

Prince, V. E. (2002). The Hox Paradox: More Complex(es) Than Imagined. Developmental Biology 249(1): 1-15.

Steller, H., Fischbach, K. F. and Rubin, G. M. (1987). Disconnected: a locus required for neuronal pathway formation in the visual system of Drosophila. Cell 50(7): 1139-1153.

CHAPTER TWO

Insect

Genes and Development—

Insect Hom-C Genes and Development—Lessons from Drosophila and

Beyond

Lisa K. Robertson and James W. Mahaffey

Department of Genetics North Carolina State University Campus Box 7614

Email: [email protected] [email protected] Office Phone: 919-515-5791

Lab Phone: 919-515-5815 Fax: 919-515-3355

THE DROSOPHILA HOM-C/HOX GENES

The word homeotic has its origins in the Greek word, homoios, meaning similar. William Bateson coined the term “homeosis” in 1894 to describe a mutation that transforms one structure to the likeness of another, particularly with respect to repeated systems such as segments, petals/sepals/stamens, vertebrae, etc. In Drosophila genetics, the term is used to describe a mutation that causes one segment to lose its normal identity and take on the appearance of another segment. There are many homeotic mutations scattered throughout the Drosophila genome, but two regions on the right arm of the third chromosome are unique. These two clusters of homeotic genes, the Antennapedia Complex (Ant-C) and the Bithorax Complex (Bx-C), are referred to jointly as the Homeotic Complex (Hom-C) or Hox complex (Figure 1). The Hom-C genes of the Drosophila Ant-C are labial (lab), proboscipedia (pb), Deformed (Dfd), Sex combs reduced (Scr), and Antennapedia (Antp) 1. The Hom-C genes of the Drosophila Bx-C are Ultrabithorax (Ubx), abdominal-A (abd-abdominal-A), and Abdominal-B (Abd-B). The genomic organization and the regional expression patterns of these genes are co-linear in the Drosophila embryo in general, a gene’s domain of expression along the anterior-posterior body axis correlates with its genomic position within the complexes.

The Hom-C/Hox genes act to direct specific segment identities along the insect anterior/posterior body axis, and the loss or misexpression of a Hom-C/Hox gene is often sufficient to transform whole body regions into a different segmental identity. Two examples of such homeotic transformations in Drosophila are shown in Figure 2. Pictured is a gain of function Drosophila Antennapedia (Antp) mutant, in which the antennae are transformed to legs. Also pictured is a loss of function Antp end stage larva, exhibiting a transformation of the second and third thoracic segments towards a first identity. Without doubt, Drosophila melanogaster is the most studied organism with regards to the Hom-C genes. These powerful transcription factors were initially identified and characterized in the fruit fly, and much of what is known of their structure, regulation, and function has come from these studies.

1 Note that Drosophila gene names follow a standard nomenclature—genes are generally named for the mutant phenotype, and

Figure 2: Phenotypic consequences of Insect Hom-C Mutations

Both loss function mutations (D, E) and gain of function mutations (B) in the Hom-C genes can have dramatic phenotypic effects in Tribolium and Drosophila. A scanning electron micrograph of a wild type (A) and a mutant adult Drosophila head (B). The ectopic activation of Antp in the antennae (ANT) causes transformation to a second thoracic leg (HOM LEG). In a wild-type first instar Drosophila larva (C), the three thoracic segments differentiate characteristic denticle

patterns. The first thoracic denticles include a main belt (marked with an arrow), and a second, small patch of denticles called the “beard” (marked by an asterisk). The morphology of the first thoracic denticles differs from those of the second and third segments, with the latter being smaller and finer. In the cuticle of an end stage larva lacking Antp (D), the second and third thoracic segments are transformed towards a first thoracic identity. Note the appearance of a beard in the second thoracic denticle belt (marked by an asterisk), and the morphological change of the second and third thoracic denticles to a first thoracic type (arrows). Fluorescent,

deconvoluted images of a wild type (E) and a TcDfd1 /TcDfd1 mutant Tribolium larva (F). Normal

antennae (ANT), mandibles (MN), maxillary palps (MAX), and labials palps (Lab) are visible in the wild type larva. In a larva lacking TcDfd function (F), the mandibles are transformed to homeotic antennae (HOM ANT). Photo credits: Tribolium images courtesy of S.J. Brown. Drosophila head images courtesy of Fly Base (Drysdale et al. 2005):

The HOX complexes are conserved in virtually all animals, as are the axial patterning functions and co-linear relationship between expression patterns and genomic arrangement. The Coleopteran beetle, Tribolium castaneum (red flour beetle) represents the best-characterized non-Drosophilid insect species. Numerous studies have characterized mutant phenotypes, expression patterns, and the genomic structure of the Tribolium HOM-C genes, making this insect an excellent model for comparative studies with Drosophila. The two are estimated to be approximately 300 million years diverged from their last common ancestor (Brown et al. 1999b). As in most animal species (other than Drosophila), the Tribolium HOM-C genes are clustered together in a single complex, differing from the split complexes observed in several Drosophila species (Beeman 1987 Ferrier and Akam 1996 Davenport et al. 2000 Powers et al. 2000 Cook et al. 2001 Negre et al. 2003). The size of the HOM-C clusters varies between the two insect species. For example, the portion of the Tribolium complex corresponding to the Drosophila ANT-C is approximately 279 kb, while the same region in Drosophila includes approximately 439 kb (Brown et al. 2002).

Figure 3: Regional Hom-C gene expression in Drosophila embryos

Figure 4: Segments, Parasegments, and Compartments

to the ectoderm. The domains of HOM-C gene expression can overlap in the ectoderm, but their expression domains in the visceral mesoderm are mutually exclusive (Treisman et al. 1989). As noted in the discussion below, HOM-C gene expression within the CNS is also frequently shifted, although by only half of a segment, such that the expression boundaries are parasegmental. Dfd is an exception, with its expression boundaries shifting anteriorly in both the visceral mesoderm and the CNS (Martinez-Arias et al. 1987).

The Embryonic Expression and Mutant Phenotypes of Deformed and Sex Combs Reduced in Drosophila and Tribolium

Two Hom-C genes are central to this study—Deformed (Dfd) and Sex Combs Reduced (Scr). Their patterns of expression and mutant phenotypes are well characterized in both Drosophila and Tribolium, although the studies in the fruit fly are more comprehensive. Following is a brief summary of the current knowledge regarding Dfd and Scr in each of these organisms.

Dfd is required for proper development of the Drosophila mandibular and maxillary segments (Wakimoto et al. 1984), as well as proper cell fate specification within the embryonic brain (Hirth et al. 1998 Hirth et al. 2001). Work by Wakimoto and Kaufman (1984) identified the requirement for Dfd in specification of the embryonic head. The Dfd gene encompasses about 11 Kb, having five exons, which produce a single transcript of 2.8 kb and a protein of 586 amino acids (Regulski et al. 1987). Dfd is the first Hom-C gene to be expressed, with protein first detectable at the cellular blastoderm stage (Mahaffey et al. 1989). Early embryonic expression of Dfd encircles the embryo, encompassing the cephalic furrow. Later, accumulation is detected transiently in the primordia of the hypopharyngial lobes, and in the developing mandibular and maxillary segment. Expression in the CNS and mesoderm is offset anteriorly, so that mesodermal and neural expression is parasegmental. Dfd expression is also present in cells between the developing dorsal ridge and the optic lobe. By late germ band contraction, Dfd is strongly expressed in cells within the frontal sac that will contribute to the larval eye/antennal discs.

and fail to internalize during head involution2. Lost or reduced cephalopharyngeal structures include the mouth hooks, H-piece, maxillary cirri, maxillary sense organs, and portions of the antennal-maxillary sensory complex. Figure 5 shows a graphic representation of the Drosophila larval head and its cephalopharyngeal structures.

A Dfd homologue, TcDfd, was cloned and characterized in Tribolium (Brown et al. 1999a). Like Drosophila, TcDfd is the first Hom-C gene expressed in the developing embryo. In general, embryonic expression is similar to flies, being present in the mandibular and maxillary segments, their appendages and the dorsal ridge, but unlike Drosophila, TcDfd transcripts are evident in the extra-embryonic serosa. As the maxillary and mandibular appendages develop, TcDfd expression modulates, appearing weaker in the mandibular appendages.

The embryonic function of TcDfd was investigated using both genetic analysis and RNA interference (RNAi) (Brown et al. 2000). Unlike Drosophila, which exhibits no homeotic transformation in Dfd null mutants, the loss of TcDfd in Tribolium transforms the mandibular appendages to antennae and eliminates the endites from the maxillary coxopodites (Figure 2F). The depletion of a Dfd homologue, Of'Dfd, in Oncopeltus fasciatus (milkweed bug) results in the transformation of both the maxillary segment and mandibular segment towards antenna (Hughes and Kaufman 2000), and it has been suggested that the antenna may represent a ‘ground state’ for gnathal segment identity (Stuart et al. 1991). The differences observed between Drosophila Dfd mutants and loss of Dfd in other insects may, however, simply reflect the fact that there are no larval appendages in Drosophila. The phenotypes of flies and other insects may, therefore, still represent equivalent developmental alterations, though one must be careful interpreting any phenotype as “ground state”, as it is likely that some patterning genes are still present to influence appendage development.

The conservation of Dfd function across phyla was tested by ectopically expressing TcDfd in developing Drosophila embryos (Brown et al. 1999a). It was previously shown that ectopically expressed endogenous Dfd would auto-activate epidermal Dfd transcription in the

Figure 5: Structures of the Drosophila larval head

The cuticular and cephalopharyngeal structures of the Drosophila larval head and first thoracic segments. Abbreviations, followed by the segmental origin: dw—dorsal wing, Acron latp— latticed piece (dorsal bridge), Acron vb—vertical bridge, Acron es—epipharyngeal sclerite, Labral mt—median tooth, Labral anso—antennal sense organ, Antennal ppw—posterior pharyngeal wall, Intercalary lp—lateral process, Mandibular vw—ventral wing, Mandibular tr—T-ribs, Mandibular mh—mouth hook, Mandibular (Tip)/Maxillary (Base) mxso—maxillary sense organ, Maxillary dmp—dorsal-medial papilla of the mxso, Antennal dlp—dorsal-lateral papilla of the mxso, Mandibular ci—maxillary cirri, Maxillary vo—ventral organ, Maxillary ds—dental sclerite, Maxillary hs—hypostomal sclerite (H-piece), Labial db—denticle belt, Thoracic db-a—first thoracic anterior belt, Thoracic db-p—first thoracic posterior belt, Thoracic ki—keilin's organs, Thoracic bd--ventral kolbchen (black dot organs), Thoracic. (Jürgens et al., 1986, The Fly Base

ventral posterior region of the labial, trunk, and abdominal segments (Kuziora and McGinnis 1988). The TcDfd transgene, driven by a heat shock promoter, was also able to activate endogenous Dfd transcription in very similar pattern (Brown et al. 1999a). It also directed the expression of a Dfd autoregulatory element, fused to LacZ. Finally, TcDfd expression, in Drosophila embryos lacking a functional Dfd protein, was able to partially rescue the Dfd null embryonic phenotype and did so to the same extent as the ectopic expression of the endogenous gene. These studies illustrate the strong conservation of function between Hom-C genes in different species.

Spanning about 100 kb, Drosophila Scr encodes at least four separate transcripts derived from alternative splicing of four exons (Andrew 1995). A large portion of the Scr regulatory region is separated from the coding region by the ftz segmentation gene (Figure 1). Mutations effecting embryonic function are 5’ to the ftz gene, closer to the Scr transcription unit, while semi-lethal lesions map 3’ of the ftz gene, towards Antp (Pattatucci and Kaufman 1991 Pattatucci et al. 1991). The Scr protein product is 417 amino acids with a molecular weight of 44 kDa (Andrew 1995).

Scr is a particularly interesting gene, in that it controls segment identity in two different tagma of the Drosophila embryo—the labial gnathal segment and the adjoining first thoracic trunk segment. Scr transcripts are first detected as gastrulation begins, and continue to be present throughout the remainder of embryogenesis (Martinez-Arias et al. 1987). Though lagging 2-3 hours behind RNA localization, protein accumulation follows the pattern of RNA transcripts, being found throughout the labial lobes and the anterior half of the first thoracic segment (Mahaffey and Kaufman 1987). In addition, a few cells in the dorsal ridge also express the Scr protein. As the germ band begins to contract, Scr protein accumulates in the midgut visceral mesoderm (Tremml and Bienz 1989 Reuter and Scott 1990). Scr also accumulates in the embryonic salivary gland primordia prior to their invagination, however, Scr is entirely absent after invagination of the developing glands (Reuter and Scott 1990).

genes, there is no larval homeotic transformation of the labial segment corresponding to the absence of Scr. Loss of Scr eliminates structures generated by the labial segment—the salivary glands and the bridge of the H-piece (Panzer et al. 1992).

Both the expression and function of the Scr homologue appear conserved in Tribolium (Curtis et al. 2001 DeCamillis et al. 2001). The Cephalothorax (Cx) gene is approximately 22 Kb in size and contains three exons. Cx is flanked by TcDfd and TcFtz, as in Drosophila, and it produces a single 1.9 Kb transcript. Cx is expressed as the germ band rudiment condenses, in the ectoderm of parasegment 2 (posterior maxillary/anterior labial). As development continues, Cx transcripts accumulate strongly in the mesoderm of the first thoracic segment, and to a lesser degree, in the mesoderm of the second and third thoracic segments. As the appendages begin to form, Cx is expressed in the ectoderm of the labial appendage primordia. Finally, when the germ band is fully retracted, Cx transcripts are present at the base of the labial appendages, in a few cells in the in the mandibular mesoderm, in a few cells in the posterior compartment of the maxillary limb, the anterior dorsal portion of the first thoracic segment, and in a segmentally repeated pattern in the developing central nervous system (Curtis et al. 2001).

Genetic analysis and RNAi was used to study the role of the Scr homologue, Cephalothorax (Cx), in Tribolium. There are two classes of mutations in the Cx gene of Tribolium (Beeman et al. 1989 Curtis et al. 2001). RNAi was used to determine which is likely to be the null phenotype, and this is a transformation of the labium to antennae and a fusion of the first thoracic segment to the head (Curtis et al. 2001). Interestingly, the labium to antennae transformation is observed in Drosophila adults only when both pb and Scr are removed (Percival-Smith et al. 1997). The second class of Cx phenotype is more complicated, causing the labium to acquire a maxillary identity, and a duplication of the first thoracic segment (Beeman et al. 1989). The reason for this transformation is yet unclear.

expression are lethal and accompanied by gross morphological defects, and even when viable adults are produced, fitness is may be greatly effected. Therefore, strict spatial and temporal control of the Hom-C selector genes is critical to normal body patterning.

The Hom-C genes’ clustered genomic structure and co-linear expression are rather unique in their strong conservation across phyla. This raises the question—why are the Hom-C/Hox genes found in such highly conserved arrangements, and is this a requirement for their expression and function? Studies in vertebrates reveal shared Hox regulatory regions and indicate that changes in the arrangement of the complex can impact the expression of multiple Hox genes (Gould et al. 1997 Sharpe et al. 1998). That splits in the Hom-C/Hox cluster have only been observed (thus far) in (Von Allmen et al. 1996 Adams et al. 2000 Negre et al. 2003 Negre et al. 2005) and in the nematode, Caenorhabditis elegans (Ruvkun and Hobert 1998), suggests a strong evolutionary constraint for the ordered Hom-C/Hox cluster, and perhaps a loosening of this constraint in these two organisms, in which shared Hom-C gene regulatory regions have not been identified. The conservation of ordered, clustered genomic organization may be required for the temporal and spatial co-linearity observed in Hom-C expression (van der Hoeven et al. 1996 Kondo and Duboule 1999), although the mechanism underlying this requirement has yet to be determined (Kmita and Duboule 2003). In spite of these questions, some mechanisms of Hom-C gene regulation in Drosophila are understood.

Initiation of Dfd and Scr expression in Drosophila

In the Drosophila embryo, the establishment of early Hom-C expression occurs during the syncitial or early cellular blastoderm stage and is controlled by the early upstream gap and segmentation gene products. Several studies have demonstrated regulation of Dfd and Scr expression by these early acting gene products and other factors.

Early Dfd activity is not significantly affected by changes in the zygotic gap genes, but mutations in the pair-rule genes do affect Dfd expression (Jack et al. 1988). Mutations in hairy, runt, even-skipped (eve), fushi tarazu (ftz), paired (prd), odd-paired (opa), odd-skipped (odd), and engrailed (en) all effect Dfd expression. odd and ftz mutants allow a spatial expansion of Dfd expression, indicating negative regulation of Dfd by these gene products. The remaining mutants contract Dfd expression by varying degrees, indicating a role in the early activation of Dfd. Some gap gene input into Dfd regulation has been reported. Embryos lacking the gap gene, tailless (tll), exhibit a transient ectopic patch of Dfd, just anterior to its normal expression domain in the cephalic furrow anlagen, and this patch is though to contribute to the enlargement of the dorsal ridge at the expense of the optic lobe (Reinitz and Levine 1990). This may occur through the action of ftz, which is also mis-expressed in tll mutants. Jack and McGinnis explored the combinatorial effects of the maternal effect gene bicoid (bcd), the gap gene hunchback (hb), and the pair-rule gene eve, on Dfd expression (Jack and McGinnis 1990). In their model, the Dfd regulatory sequence integrates three levels of the embryonic segmentation hierarchy—an intermediate concentration of the maternal gene, bcd in combination with the gap gene, hb and the pair rule genes eve and ftz—to achieve the appropriate activation and positioning of Dfd. Each of these genes encode transcription factors, thus their effect on Dfd transcription may be direct.

been proposed as possible members of a primordial Hox gene cluster (McGinnis and Krumlauf 1992 Macias and Morata 1996 Gallitano-Mendel and Finkelstein 1998).

Previous work suggests a combinatorial role for early patterning genes in the initiation of Scr expression. The pair-rule segmentation gene, ftz, was shown to be required for the initial activation Scr (Riley et al. 1987), and the loss of both maternal and zygotic hb expression affects the spatial placement of Scr expression. More recent work demonstrates a direct role for Hb in the activation of Antp, and the authors propose that this might also be true for the gap protein, Giant (GT) in the activation of Scr (Wu et al. 2001).

Maintenance of Dfd and Scr Expression in Drosophila

The maintenance of Drosophila Hom-C gene expression or repression is required beyond the early stages of embryogenesis, after the early patterning gene products have dissipated. Because cells continue to express the Hom-C genes in spatially restricted patterns throughout development, cells must “remember” their Hom-C transcription state to avoid inappropriate activation or loss of selector gene function. This long-term retention of Hom-C transcription state is taken over by two large groups of trans-acting proteins—the Polycomb Group (PcG) (Simon 1995) maintains the Hom-C genes in a repressed state, while the Trithorax Group (TRX) (Ingham 1998) maintains the genes in an active state. Both the TRX group and PcG group proteins are widely conserved between Drosophila and mammals (Muller et al. 1995 Simon 1995).

formation of higher order chromatin structure associated with histone methylation (Cao et al. 2002). The PREs do not have a specific recognizable consensus DNA sequence (Jacobs and van Lohuizen 1999), and the PcG proteins are likely recruited to the PREs by either the repressive gap or pair-rule proteins, or another DNA binding intermediary.

The Drosophila Trithorax (TRX) group proteins serve a reciprocal role to the PcG proteins and maintain the active transcriptional state of the Ant-C and the Bx-C genes (Breen and Harte 1993 Kennison 1993 Gindhart and Kaufman 1995). Trithorax (TRX), the namesake and best-characterized member of the group (Ingham and Whittle 1980 Ingham 1985), accumulates in a spatially modulated pattern, beginning with pair-rule-like stripes in the posterior of the embryo. This may regulate early Bx-C transcription (Sedkov et al. 1994), as trx mutants have altered Bx-C gene expression during germ band extension. However, Ant-C gene expression is unaltered until late stages of development. There are alleles of trx that affect late Ant-C gene expression with no effect upon Bx-C expression, indicating the presence of multiple products, with multiple roles, from the trx locus. Null mutants of trx mimic loss of function mutations in the Ant-C and Bx-C clusters, sometimes producing flies with six dorsal thoracic appendages of wing-like identity, the phenotype for which the gene and the group are named (Ingham 1998). Like the PcG proteins, it is suspected that the TRX group may not bind DNA directly, but act through other proteins with DNA binding activity. Interestingly, analysis of PcG and TRX protein localization on salivary gland polytene chromosomes demonstrates that these proteins bind many of the same chromosomal regions, which include the Ant-C and the Bx-C. Thus, interaction between these opposing regulatory proteins at the same chromosomal sites may be important to their function (Kuzin et al. 1994 Chinwalla et al. 1995). An important final note—although these protein complexes function to maintain the initial C gene activity state in a particular region, the Hom-C genes are by no means “locked in” to a particular activity state for the duration of development. This is evident in the dynamic patterns of expression observed for individual Hom-C genes during the course of development and discussed earlier in this chapter.

reduced eye and head structures. Some of the defects resemble the head defects observed in hypermorphic alleles of Dfd, and the observed mxc phenotypes were shown to be dependent upon Dfd dosage. In the absence of mxc, Dfd is ectopically expressed in the eye/antennal disc. Taken together, the results support mxc as a regulator of Dfd expression, required to limit Dfd expression in the eye/antennal disc. Two Trx loci were identified in a screen for mutations that enhanced the semi-lethality of hypomorphic Dfd alleles-- kismet (kis) and snaggletooth (snt) (Gellon et al. 1997). Both are suppressors of Pc function and cause embryonic head defect when homozygous. Interestingly, kis was originally identified in a screen for modifiers of Pc and Antp mutations, revealing that this gene function in maintaining the expression of multiple Hom-C genes.

Two specific regions of regulatory sequence, one 15 Kb downstream and one 40 Kb upstream of the Scr transcription start site, were found to respond genetically to PcG and Trx loci to effect Scr expression during later embryogenesis and imaginal development (Gindhart and Kaufman 1995). A 10 Kb sequence located in the 40 Kb region was able to repress the expression of an Scr reporter construct in a PcG and Trx dependent manner. Two PcG genes, extra sex combs (esc) and Additional sex combs (Asc) are required for the silencing of Scr in the second and third thoracic leg discs (Pattatucci and Kaufman 1991). A Trx loci, Brahma, was shown to be required for the maintenance of proper Scr expression in the first thoracic imaginal leg disc (Tamkun et al. 1992). Dorsal Switch Protein 1 (DSP1) is a Drosophila HMG protein that can act as a PcG group or a Trx group protein, depending upon the target locus (Decoville et al. 2001). DSP1 was shown to be specifically involved in the spatio-temporal regulation of Scr expression in the first thoracic imaginal discs (Rappailles et al. 2004) The authors were able to localize two 1 kilobase regions in the 10 kilobase regulatory sequence, previously identified by Gindhart and Kaufman, that interact directly with DSP1. The trx gene product has a role in maintaining both Dfd and Scr expression, along with a number of other Hom-C genes (Breen and Harte 1993). trithorax (trx) mutants exhibit reduce, but not absent, expression of Dfd and Scr—the reduction in expression is insufficient to cause significant phenotypic effects, suggesting that other factors are also at work to maintain Dfd and Scr during late embryogenesis.

of PS2 during mid-embryogenesis (Henderson and Andrew 2000). Interestingly, this is a feedback loop in which the maintenance of Scr expression by hth leads to the inhibition of hth expression by Scr and ultimately the loss of Scr expression in the salivary gland primordia.

Post-translational Regulation of Scr

Not all Hom-C gene regulation occurs at the level of transcription Hom-C protein function is also modulated post-translationally. The N-terminal arm of the homeodomain is important to the specificity of homeodomain function (discussed below), and this region of the Scr protein is a target for a regulatory subunit of the serine-threonine-specific protein phosphatase 2A (dPP2A, B’) (Berry and Gehring 2000). The phosphorylation state of two key residues in the N-terminal portion of the homeodomain (Thr6 and Ser7) is critical to Scr function. These residues can be phosphorylated by cAMP-dependent kinase A (PKA), and dephosphorylated by dPP2A in cultured cells. The ectopic expression of an Scr protein construct mimicking constitutive dephosphorylation (with Thr6 and Ser7 altered to alanine), reproduces the embryonic phenotype induced by ectopic expression of the wild-type protein (loss of some mouthpart structures, ectopic salivary gland formation, and transformation of the second and third thoracic segments towards a first thoracic identity). Conversely, an ectopically expressed Scr protein construct, in which the Thr6 and Ser7 residues are altered to mimic constitutive phosphorylation, is unable to reproduce the embryonic ectopic Scr phenotype and has a reduced ability to bind a target DNA sequence in vitro. Further, dPP2A,B' is appears to be required for endogenous Scr function, as embryos lacking dPP2A,B' fail to form salivary glands.

epidermal expression in the posterior compartments of the thoracic and abdominal segments (Pelaz et al. 1993). Interestingly, this expansion is dependent upon moderate levels of ectopic Antp (normally repressed by the Bx-C proteins), which activates Scr in the context of the posterior trunk segments in the absence of the Bx-C proteins.

In two recent studies, Miller and collaborators utilized the UAS-Gal4 system (Brand and Perrimon 1993) to investigate the embryonic cross-regulatory relationships of the Hom-C genes (Miller et al. 2001a Miller et al. 2001b). The Hom-C genes demonstrate tissue-specific cross-regulatory activities. For example, ectopic Scr and ectopic Dfd activate pb in the ectoderm, which is consistent with the normal role for these proteins in activating pb expression in the embryo (Rusch and Kaufman 2000). In the mesoderm, however, ectopic Scr represses the expression of pb (Miller et al. 2001a). The hierarchy of cross-regulation in the CNS also differs from that in the epidermis. Ectopic Ubx and Abd-A are only able to repress two Hom-C genes (Scr, lab) in the CNS, whereas Ubx and Abd-A have greater cross-regulatory effects in the epidermis. Miller et al. conclude that no simple model comprehensively describes the regulatory relationships between the Hom-C genes. Their interactions in the CNS and mesoderm are more complex, and signaling cascades likely contribute to this complexity. Assessment of the relationship between cell autonomous regulation and signal transduction led them to propose a new hierarchical relationship in the mesoderm, called “autonomic dominance”, in which the extrinsic determination of cell fate via signaling can be overridden by the expression of Hom-C proteins.

Several regionally specific Dfd autoregulatory elements have been characterized which maintain Dfd expression in a tissue specific manner (Kuziora and McGinnis 1988 Bergson and McGinnis 1990 Regulski et al. 1991 Zeng et al. 1994). These elements may be located upstream of the promoter or within intronic regions. For example, a CNS-specific autoregulatory element of Dfd maps within the large Dfd intron (Lou et al. 1995), while an epidermal enhancer resides upstream of the Dfd promoter (Bergson and McGinnis 1990 Zeng et al. 1994).

identity, as evidence by the partial loss of mouthpart structures and the formation of ectopic salivary glands (Zeng et al. 1993). Thus, Scr overrides the normal patterning directed by Dfd in the mandibular and maxillary lobes. This could be due to a higher level of Scr than Dfd in these regions. However, there is no phenotypic effect in the trunk posterior to the second thoracic segment. Ectopic Scr appears unable to supersede the functions of endogenous Ubx, Abd-A, and Abd-B, even though these proteins are likely less abundant. Similarly, the ubiquitous ectopic expression of Ubx causes the transformation of segments anterior to the first abdominal, including the development of abdominal type denticles in the head and in the thorax (Gonzalez-Reyes and Morata 1990), but those segments posterior to the first abdominal are not altered. Therefore, like Scr, Ubx is unable to override Abd-A and Abd-B function in the epidermis. There are, however, violations to the posterior prevalence model. For example, the ectopic expression of Dfd is able to induce ectopic mouth hooks in the labial segment, alter the denticle patterning of the second thoracic segment to resemble the first thoracic segment, and produce ectopic sensory cirri and sclerotized mouthpart-like material in the trunk segments (Kuziora and McGinnis 1988) thereby apparently altering the patterns directed by the endogenous Hom-C proteins in these cells.

THE HOM-C PROTEIN HOMEODOMAIN

example, Dfd, and a mammalian homologue HoxD-4 (Hox-4.2), are conserved within the homeodomain at 55 of 60 residues. HoxD-4 expression in Drosophila is able to activate a Dfd responsive reporter in embryos, and produces adult phenotypes similar to those caused by a dominant gain of function Dfd allele (McGinnis et al. 1990).

In addition to the Hom-C genes, there are numerous other Drosophila homeodomain encoding genes. Several are well studied and provide much of the basis for our understanding of homeodomain structure and function. Particularly well-characterized examples include the segmentation genes engrailed (en) (Desplan et al. 1985 Desplan et al. 1988 Kissinger et al. 1990 Ohkuma et al. 1990 Jaynes and O'Farrell 1991 Ades and Sauer 1994 Clarke et al. 1994 Bourbon et al. 1995) and fushi tarazu (ftz) (Nelson and Laughon 1990 Percival-Smith et al. 1990 Florence et al. 1991 Walter et al. 1994), and the maternal factor bicoid (bcd) (Hanes and Brent 1989 Treisman et al. 1989 Hanes et al. 1994). All of the Hom-C proteins, and other homeodomain containing proteins including Even-skipped (Eve), Ftz and En, have a conserved glutamine (Q) residue at position 50 of the homeodomain and are referred to as the Q50 homeoproteins (Biggin and McGinnis 1997).

conclusions from these studies can be reasonably applied to the remaining Hom-C genes and probably other Q50 homeodomain proteins, as well.

The studies undertaken to characterize homeodomain function included genetic interactions, transcription assays, DNA/protein cross-linking assays, immunoprecipitation, and DNA foot printing experiments. The results of this work characterized the homeodomain as a DNA binding motif and the Hom-C proteins as transcriptional regulators (Desplan et al. 1985 Müller et al. 1988 Krasnow et al. 1989 Samson et al. 1989 Johnson et al. 1995). In fact, the homeodomain is one of the most common DNA binding motifs found in eukaryotic transcriptional regulators, second only to the C2H2 zinc finger class (Tupler et al. 2001).

Numerous in vitro DNA binding studies with the Hom-C homeodomains resulted in the characterization of a core consensus DNA recognition sequence. An early study of Hom-C protein/DNA interactions used the Antp homeodomain (Antp-HD) recognition sequence, which exhibited specific DNA binding affinity for the sequence 5’-TAATG-3’ (Müller et al. 1988 Affolter et al. 1990 Ekker et al. 1991). Other studies of Hom-C/DNA interactions involve partial or full-length Dfd, Antp, Ubx, and Abd-B proteins (Desplan et al. 1988 Müller et al. 1988 Affolter et al. 1990 Ekker et al. 1991 Ekker et al. 1992 Ekker et al. 1994 Wilson et al. 1995). Although Abd-B appears to bind preferentially to a slightly different core sequence of 5’-TTAT-3’ (Ekker et al. 1994), the Hom-C protein homeodomains consistently recognize a consensus core sequence of 5’-TAAT-3’ within characterized recognition sites of 5-9 base pairs.

Figure 6: The Hom-C protein homeodomain

although the Hom-C proteins recognize and bind specific DNA sequences, it appears that they can overlap, and that all recognize the same core sequence. The conclusion is that there must be additional in vivo mechanisms involved in directing their biological patterning specificities.

Functional Specificity of the Hom-C Homeodomain

Because of the strong amino acid conservation and apparent similarities in target site recognition, several studies have attempted to dissect Hom-C protein homeodomains to identify specific residues critical to their biological function. Several groups approached this problem by constructing chimeric Hom-C proteins (swapping various regions of one Hom-C protein for another), then expressing the chimeras in Drosophila embryos and adults to assay developmental outcomes and the effects on target gene expression. Although these types of studies have involved the majority of the Hom-C proteins, I summarize, below, the results of selected swapping studies involving the Ubx homeodomain, inserted into the Dfd protein (Kuziora and McGinnis 1989 Dessain et al. 1992 Lin and McGinnis 1992) and the Scr homeodomain, inserted into the Antp protein (Gibson et al. 1990 Furukubo-Tokunaga et al. 1993).

Substitution of the Ubx homeodomain into an otherwise normal Dfd protein is sufficient to introduce Ubx target regulation, and override normal Dfd activity. Wild type Dfd can autoregulate its own expression in some cells, binding to defined autoregulatory elements to maintain its transcription (Kuziora and McGinnis 1988 Lou et al. 1995). The ectopic activation of wild type Dfd results in the development of ectopic mouthpart material and maxillary cirri in the trunk segments, and the autoactivation of Dfd outside of its normal domain. Activation of a chimeric heat shock inducible Dfd construct, encoding the Ubx homeodomain in place of the Dfd homeodomain, results in a transformation of the larval head segments toward a thoracic identity, as also results from ectopic Ubx expression, (Gonzalez-Reyes and Morata 1990). The ectopic activation of Antp P1, a known Ubx target (Beachy et al. 1988), and the concomitant loss of Dfd autoregulation are also observed (Kuziora and McGinnis 1989). Thus, the Dfd/Ubx chimera, with only the homeodomain region changed, alters the protein function to resemble a wild type Ubx protein construct.

an otherwise normal Dfd protein construct (Lin and McGinnis 1992). Each construct was tested for its ability to activate the Dfd autoregulatory element and the Antp P1 promoter. Constructs retaining the Ubx N-terminal homeodomain region also retain the larval head to thoracic transformation produced by the original Dfd/Ubx construct and activate ectopic Antp P1 transcription. Further, substituting only a few N-terminal Ubx homeodomain residues (numbers 0-9, see Figure 9) is sufficient to alter ectopic Dfd function and allow initiation of ectopic Antp P1 transcription. A chimera containing the Ubx homeodomain with the N-terminal Dfd homeodomain region fails to activate ectopic Antp P1 transcription and loses the ability to induce a head to thorax transformation. Both the N-terminal and C-terminal homeodomain regions are required for full Dfd autoactivation—no autoactivation is possible without the N-terminal Dfd region and only weak autoactivation is possible with the C-terminal portion absent. In summary, these studies demonstrate that the homeodomain is capable of directing Hom-C protein function, and specifically, the N-terminal homeodomain region alone is sufficient to alter Hom-C protein function. The C-terminal flanking region also contributes to functional specificity.

Others took a similar approach in comparing two Ant-C proteins, Antp and Scr. Gibson, et al., (1990) used the degree of transformation and target gene transcript accumulation to assess the functional specificity of a series of ectopically expressed, chimeric proteins. The Scr homeodomain was substituted into an otherwise normal Antp protein construct and the chimera was expressed in embryos and adults, utilizing an inducible heat shock promoter. This work determined that functional specificity lies in residues both within and adjacent to the homeodomain, and corroborated results obtained with the Dfd and Ubx studies.

the chimeric Scr/Antp protein (wild-type Antp with the Scr homeodomain), a functional Scr protein is reverted to a functional Antp protein. Therefore, substitution of these four amino acids alone is sufficient to determine the functional difference between Antp and Scr. Similar studies involving Ubx and Antp, and Dfd and Abd-B produced similar results (Kuziora and McGinnis 1990 Chan and Mann 1993 Zhu and Kuziora 1996). Therefore, functional specificity between the homeodomains of two different Ant-C proteins appears to be determined by only a few critical amino acids within the N-terminal portion of the homeodomain, while residues within the large region N-terminal to the homeodomain directs only the extent of the protein’s functional effects.

As discussed previously, the phosphorylation state of two N-terminal homeodomain residues (6 and 7) is critical to Scr function. At position 6 of the homeodomain, Scr has a threonine residue, while Antp has a non-phosphorylatable glutamine residue. The phosphorylation state of this N-terminal homeodomain residue may, at least in part, explain some of the differing functional specificities of the Scr and Antp homeodomains.

differences in target DNA sequence recognition, as the homeodomains are unchanged between isoforms. This seeming incongruence between the specificity of Hom-C protein function and the apparent lack of specificity in DNA target recognition is frequently referred to as the “HOX Paradox.” It seems likely then, that functional specificity results from spatial and/or temporal differences in expression and differential protein interactions. The N-terminal region of the homeodomain and the highly variable regions outside of the conserved portion of the homeodomain may guide co-factor interactions to direct the developmental specificity brought about by the homeotic selector genes.

COOPERATIVE INTERACTIONS BETWEEN HOM-C PROTEINS AND OTHER FACTORS

This is nicely illustrated by studies utilizing the Hom-C protein Dfd and a known Dfd autoregulatory sequence. Multiples of a small Dfd binding fragment, derived from a Dfd autoactivation enhancer, is sufficient to drive β-galactosidase reporter expression in the maxillary segment of Drosophila embryos. However, the loss of an imperfect inverted repeat sequence removes the maxillary reporter expression (Zeng et al. 1994), suggesting that other proteins must also bind to this fragment to allow the activation of transcription. Further, the Dfd protein binds strongly, in vitro, to a tandem repeat of its consensus recognition site, but can only significantly activate transcription of a reporter construct containing this site when it is accompanied by the constitutive VP16 activation domain (Li et al. 1999a). Thus, in spite of strong binding, Dfd alone is insufficient to activate the reporter without the addition of an activation sequence, further indicating a requirement for additional factors. Finally, the Dfd-VP16 fusion activates ectopic expression of Scr, in vivo, where as the wild-type Dfd protein does not regulate Scr. This demonstrates that the Hom-C proteins likely bind, in vivo, to sites that are not at normal target gene promoters or enhancers.

co-activators (Phillips and Luisi 2000). Yeast also provides an example of a homeodomain/zinc finger interaction between the PHO2 and SWI5 proteins during regulation of the HO endonuclease gene (Brazas and Stillman 1993 Brazas et al. 1995). Finally, there are homeodomain-containing proteins, which also contain additional DNA binding motifs to guide their binding specificity.

For the Hom-C proteins to direct the specification of segment identities, it is apparent additional factors or mechanisms are required beyond the simple binding of a Hom-C protein to a recognition site. How might co-factors contribute to target gene selection? We mention above two models proposed by Biggin and McGinnis (1997) for the role of co-factors during Hom-C protein target regulation. Though there is evidence for both models, the Widespread-Binding Model appears to have the most support. However, we must point out that this is not likely to be a simple process involving one, two, or even three factors. More likely, the process of target gene selection involves multiple factors, with perhaps some, like the gap gene products, expressed prior to the Hom-C proteins. Further, when thinking of co-factor function, one might expect direct interaction with the Hom-C proteins, but this need not be the case. Some proteins functioning as co-factors may not even be DNA binding proteins, but might serve as “linkers,” required to stabilize interactions between the basal transcription machinery and the Hom-C proteins. Clearly, much more investigation is needed, and because there are differences between different insect species, perhaps a comparative study can yield answers to some of these questions.

There are several genetic and biochemical characteristics we might expect in a Hom-C co-factor:

• Absence of the co-factor would duplicate full or partial Hom-C mutant phenotypes.

• A co-factor would act in parallel to the Hom-C protein, functioning neither as an upstream regulator, nor a downstream target (although we might expect some regulatory “cross talk” between the two).

• Hom-C target gene expression would be reduced or absent in flies lacking the co-factor.

Several Drosophila gene products have been described, which meet many, or all, of these criteria. The homeodomain encoding gene extradenticle (exd), is the best-characterized Hom-C co-factor (Peifer and Wieschaus 1990). Other genes proposed to encode Hom-C co-co-factors include the bZip family member cap’n’collar (cnc), two genes encoding putative novel transcription regulators, lines (lin), and apontic (apt), and the zinc finger encoding genes teashirt (tsh), disconnected (disco), disco-related (disco-r) and perhaps, buttonhead (btd).

Extradenticle (Exd) is a Hom-C Protein Co-factor

extradenticle encodes a 378 amino acid protein required for proper embryonic and adult patterning. exd is homologous to the pbx group of vertebrate genes (Rauskolb et al. 1993) and is a member of the highly conserved PbX class of homeodomain proteins (Bürglin 1994). The Exd homeodomain diverges from those of the Hom-C genes, having three extra residues between the first and second alpha helix. Atypical homeodomains of this type are referred to as three-amino acid loop extension (TALE) homeodomains. A second conserved domain is the PbC-A domain, through which Exd interacts with a second homeodomain containing protein, Homothorax (Hth) (Ryoo et al. 1999). This interaction is discussed in detail below.

Embryos lacking zygotic Exd function die during late embryogenesis and have defects in overall pattern formation (Peifer and Wieschaus 1990). Development of some head structures is disrupted, and the thoracic segments are altered, with the second thoracic segment adopting a first thoracic-like denticle pattern, and the third thoracic segment showing characteristics of both the first thoracic and second abdominal segments. Identities of the abdominal segments are transformed to those two to three segments more posterior. For example, the first abdominal segment resembles the third or fourth. No effect is detected in the posterior most segments. Weaker alleles are post-embryonic lethals often forming pharate adults with cuticle defects, and clonal analysis demonstrates that lack of Exd also causes transformation of adult structures (Gonzalez-Crespo and Morata 1995 Rauskolb et al. 1995). Null exd clones result in homeotic transformation of adult head structures, legs, and abdominal segments resembling the phenotypes observed in some Hom-C mutants. Antenna and arista are transformed to leg, while the head capsule is altered to resemble the dorsal thorax or notum. Abdominal clones in the first through fourth segments transform towards the fifth or six abdominal segments. Ubiquitous over-expression of exd has no affect on segment identities (Rauskolb et al. 1995). Because exd mutants appeared to have alterations in Hom-C target specificity, Peifer and Wieschaus proposed exd as a Hom-C cofactor.

To explore the interaction and contribution of the Hom-C proteins and Exd to development, embryos mutant for exd were compared to embryos singly mutant for Hom-C genes and doubly mutant for both a Hom-C gene and exd. Double mutants differed from either single mutant, indicating that the Hom-C genes continued to be active within their normal domains in the absence of exd, although they are unable to properly specify segmental identity (Peifer and Wieschaus 1990). For example, single exd mutants exhibit a transformation of the first abdominal segment to the third, the second abdominal segment to the fourth, and the third abdominal segment to the fifth. An abd-A mutant exhibits a transformation of these same segments to an A1 identity. The double exd abd-A mutant transforms all of these same segments to an A3 identity. These results further indicate that exd functions in a parallel pathway to the Hom-C genes and is required for the normal specification of segment identity by the Hom-C genes examined.


Watch the video: Γιατί κ Πως Τρελαίνεται ο Κόσμος (January 2022).