Information

Does parental conflict lead to genes combining important functions with functions only advantageous for one of the parents?


In a sitation of a mother-father conflict of interests, the mother might use epigenetics to turn off some genes only advantageous for the father's genes and not her own. I thought a logical father's contra-strategy would be to bind the functions only advantageous to him to some of the genes crucial for them both, so that the mother couldn't turn these genes off.

Is there any evidence for a similar mechanism of "taking some genes hostage" in a situation of a conflict of interests?


Does Human DNA Change With Time?

DNA technology has become incredibly popular and important in recent years due to its ability to facilitate criminal investigations, paternity tests, and archaeology records, just to name a few of its areas of application. Who knew that a small double helix could contain so much potential?! Even Johann Friedrich Miescher, the Swiss scientist who discovered DNA (from white blood cells), was unsure how this one molecule could record all the diverse information about a particular species!


INTRODUCTION

Every culture is characterized, and distinguished from other cultures, by deep-rooted and widely acknowledged ideas about how one needs to feel, think, and act as a functioning member of the culture. Cross-cultural study affirms that groups of people possess different beliefs and engage in different behaviors that may be normative in their culture but are not necessarily normative in another culture. Cultural groups thus embody particular characteristics that are deemed essential or advantageous to their members. These beliefs and behaviors tend to persist over time and constitute the valued competencies that are communicated to new members of the group. Central to a concept of culture, therefore, is the expectation that different cultural groups possess distinct beliefs and behave in unique ways with respect to their parenting. Cultural variations in parenting beliefs and behaviors are impressive, whether observed among different, say ethnic, groups in one society or across societies in different parts of the world. This article addresses the rapidly increasing research interest in cultural differences in parenting. It first takes up philosophical underpinnings, rationales, and methodological considerations central to cultural approaches to parenting, describes a cross-cultural study of parenting, and then addresses some core issues in cultural approaches to parenting, viz. universals, specifics, and the form-versus-function distinction. It concludes with an overview of social policy implications and future directions of cultural approaches to parenting.


Explanations for individual variability in attachment patterns

Environmental factors

Bowlby's attachment theory is a truly environmental theory as it has explained individual differences in attachment patterns (attachment types) by individual variations in caregivers' behaviour. In their seminal study [5], Ainsworth and colleagues found links between observed care-giving behaviour at home and characteristic behaviour patterns in the laboratory-based SSP. They found that the optimal, secure behaviour pattern could be linked to sufficient sensitive responsiveness at home. Sensitivity was conceptually distinguished from responsiveness, with sensitive responses defined as being guided by an appropriate interpretation of infants' signals and changing needs. The avoidant pattern could be related to rejecting, dismissing or neglecting responses to infants' signals, especially to those signals expressing negative emotions, while in the background of the resistant pattern, unreliable, inconsistent care was identified.

Although many studies demonstrated a significant link between early care and attachment, studies varied greatly regarding in estimates of the strength of the relationship. De Wolff and van IJzendoorn [11] reviewed 66 studies to evaluate effect sizes in relation to the methodology used for assessing caregivers' sensitivity. They showed that caregiver sensitivity has been defined and operationalised in many different ways over the preceding thirty years, but however measured, it was far from being an exclusive determinant of the quality of attachment. Indeed, sensitivity accounted for less variance than was expected (effect sizes 0.24-0.32).

Over the last three decades, it has been shown that different demographic risk factors, especially if accumulated may effect the development of attachment, presumably through their proximal or distal influence on parenting [12]. Income and family size, parental age and education, major stressful events, such as loss of a parent, birth of a sibling, severe illness, marital relationships and breakdown affect the quality of attachment relationships [13–19].

It has been expected that secure attachment is promoted by the psychological health of parents, especially mothers. Empirical studies have provided contradictory results, but the majority found that mothers depressed postnatally were more likely to develop insecure attachment relationships with their infants [20–23]. Studies in general have not been able to find direct associations of mother-infant attachment with child care arrangements and with mothers' social support systems [12], but in high social risk groups, lack of support correlated with higher rates of insecure attachment relationships [24–26], while extensive support was found to promote security [27, 28]. Security of mother-infant attachment has been found to be related to mother's mental state with respect to close (attachment) relationships. In several studies, the security of maternal attachment representations, as assessed by the Adult Attachment Interview (AAI) [29], has been found to be significantly related to mother-infant attachment security [30].

It has also been shown that while isolated individual risk factors may not have a significant effect on parent-child attachment, the accumulation of adversity may result in sub-optimal relationship development and insecurity of infant attachment [12]. Raikes and Thompson [31] have tested the effect of multiple social and economical factors on attachment and confirmed that their effects were mediated by mothers' care-giving behaviour. In several longitudinal studies, cumulative risk indices and life stress were used to explain discontinuities in attachment security through the life course [32–34].

Friedman and Boyle [35] have recently reviewed attachment-related findings from the longitudinal NICHD Study of Early Child Care (SECC) aiming to identify effects of timing, extent, quality, and type of child care experiences on children's development in a large sample, using well-controlled methodology. The study followed the development of more than one thousand children, of varied backgrounds, from birth and periodically assessed attachment to their mothers. Family SES and maternal sensitivity, especially in responding to the child's distress, predicted the quality of attachment at 15 months [36, 37] and at 36 months of age [38] confirming the previously found moderate strength of the relationship. Recurring symptoms of maternal depression across the first three years predicted higher prevalence of insecure attachment at age 36 months [39]. Analysis of effects of non-maternal care, a special focus of the NICHD study, confirmed the lack of main effects of children's age of entry, quality of care and length of time children spent in non-maternal care, but also revealed interaction effects. The study found higher rates of insecurity if a low quality of non-maternal care was combined with low maternal sensitivity and more time spent in child care. Altogether, the conclusion of the study was that, under some circumstances, early non-maternal child care can be an environmental risk factor for attachment security and compromised later development.

Environmental effects on disorganized attachment

Attachment disorganisation became a focus of developmental research when rarely occurring incoherent and contradictory infant behaviours, not fitting the Ainsworth categories, appeared to be predominant among maltreated or otherwise deprived groups of infants and young children [6, 40]. Infants whose attachment strategy collapses even under the mild stress of brief separation experienced in the Strange Situation and who show high degree of incoherence and disorganisation upon reunion with their caregivers comprise on average 15% of typical populations and as high as 50-80% of high social risk groups [8]. Early disorganised attachment also proved to be one of the rare early predictors of subsequent childhood behaviour problems [41–44] and adolescent psychopathology, such as dissociative symptoms and borderline personality disorder [45, 46]. Regarding the parenting background, Main and Hesse [47] proposed that caregivers of infants displaying disorganized attachment showed bouts of "frightened, threatening, and dissociative" behaviour in interactions with their infants, and this was confirmed by later studies [48–50]. In addition, Lyons-Ruth and colleagues [51] identified a broader spectrum of anomalous parental behaviour contributing to attachment disorganisation, including also communication errors, role confusion and extreme withdrawal. To some extent, these behaviours were also found in low-social-risk groups of mothers [43, 52–55]. Using NICHD SECC data, Campbell and colleagues [39] have shown that chronic depression combined with low maternal sensitivity is associated with a higher prevalence of disorganised attachment in 3-year-old children. Mothers' unresolved trauma or loss has been considered as a potential source of disruptions in maternal care-giving behaviour [43, 47, 50, 56]. According to a recent meta-analysis, however, anomalies of caregiver's mental state and behaviour had only low explanatory power in accounting for attachment disorganization [49]. Including 12 studies examining relations between maternal unresolved loss and trauma, anomalous parenting and disorganised attachment, moderate effect sizes were found for both links between maternal unresolved mental state and anomalous behaviour and infant disorganized attachment (r = .26 and .21, respectively), as well as for the link between mother's anomalous behaviour and infant disorganization (r = .34).

Biological contribution to individual variations in infant attachment

Newborns' biologically based capabilities of self-regulation of arousal and distress states have an immediate impact on parents. The great individual variation in these capabilities can be described by dimensions such as the infant's disposition for distress or negative emotionality, irritability and soothability [57, 58]. There is a long tradition of two different interpretations of the role of infant temperament in the formation of attachment relationships. Attachment theorists have suggested that temperament has no direct effect on the quality of attachment, since infant characteristics such as difficult temperament can be accommodated by sensitive caregivers, who can still foster secure attachment relationships [59]. Temperament researchers, on the other hand, have kept emphasizing that infant-caregiver interactions in the Strange Situation reflect the infant's temperament rather than the quality of the relationship [60]. In their extensive review, Vaughn and Bost [61] argue that temperament and attachment are separate constructs, and studies showing inter-relationships on the one hand, and independence on the other result from different conceptualisations and assessments of both. There is a body of empirical research results, which has demonstrated relations between attachment quality and infant irritability, proneness to distress or stress regulation [26, 58, 62–64]. Based on their review of literature, Mangelsdorf and Frosch [65] have suggested that effects of infant temperament on attachment may be indirect and moderated by other maternal and social variables.

Several studies have found that newborn behavioural measures are related to later secure and insecure attachment classifications [66–70]. These studies are particularly interesting because neonatal measures reflect minimal social experience. Results from the Regensburg longitudinal studies have shown that poor neonatal behavioural organisation was related to disorganised behaviour in the Strange Situation at 18 months of age [71, 72]. This, together with findings that some of the behaviours characteristic of attachment disorganisation have also been found in children with developmental disorders [73, 74] imply that biological factors such as the infant's capability to organise environmental stimuli, communicate and regulate internal states and behaviour may also contribute to the development of disorganised attachment. Summarizing the empirical evidence, Barnett and colleagues [75] suggested a two-dimensional model in which both biological vulnerability as well as adverse environment might contribute to the development of atypical (disorganised) attachment behaviour.

Twin studies of attachment security and disorganisation

Since the study of parent-child attachment was so strongly driven by the environmental theory, a quantitative behaviour-genetic approach has only recently been used to investigate heritable and environmental variance components of attachment security. Most of the existing twin studies of attachment overviewed here have not been designed to focus on attachment, therefore they are quite heterogeneous regarding children's age and the method of assessment. In the Louisville Twin Study (LTS), Finkel and Matheny [76] found a significant difference between concordances of attachment classifications of 99 monozygotic (MZ) and 108 dizygotic (DZ) two-year-old twins. Model-fitting resulted in estimates of 25% genetic and 75% non-shared environmental effects. In another study of 57 MZ and 53 same-sex DZ twins aged 3.5 years, O'Connor and Croft [77] reported results of genetic model-fitting: variance estimates due to genetic, shared and non-shared environmental factors were 14%, 32% and 53%, respectively, but the genetic effect on secure vs. insecure attachment was not significant. Comparing 57 MZ and 81 DZ twin pairs, Bokhorst and colleagues [78] found only unique environmental factors accounting for the variance in disorganised vs. organised attachment, while both shared and non-shared environmental effects accounted for the variance in secure vs. insecure attachment. In a model-fitting analysis using data from 485 twin pairs, Roisman and Fraley [79] have also emphasized the role of environment (parenting quality) in accounting for the variability in toddlers' observed secure-base behaviour. Using staff/parent-rated zygosity, a model containing only shared (C) and non-shared (E) environmental variances "was able to explain the data just as well as the full [ACE] model" (p. 835) providing a heritability estimate of 0.17. Finally, using questionnaire data of attachment disorder behaviours in a very large community sample of 13,472 twins, both twin correlations and model-fitting results suggested a strong genetic influence on attachment disorder behaviour, especially in boys [80].

It is important to recognize the power constraints of quantitative genetic studies such as those using twin comparisons. These analyses aim at assessing heritability that is, estimating how much of the population variance is due to genetic effects (the rest is environmental variance and measurement error). The twin method is based on comparison of twin correlations or concordances. If genetic effects are important for the trait in question, then correlation or concordance between monozygotic twins should be significantly greater than that between dizygotic twins. Limited sensitivity is inherent in the methodology based on detecting significant differences between twin correlations. This is not such a great limitation for many psychological and psychiatric phenotypes with substantial heritabilities of around 0.5, but may cause problems in detecting smaller, yet potentially important genetic effects. (It is estimated that any specific gene effect accounts for less than 1% of the variance of complex traits [81].) The power limitations of relatively small twin studies are not trivial. Statistical power analysis by Visscher [82] has addressed the questions of rejecting CE models within the classical twin design when the true model is ACE, i.e. there is a significant genetic variance. The simulation results have shown that necessary sample sizes become fairly large at 0.2 heritability except when shared environmental variance exceeds 0.5-0.6. Visscher and colleagues [83] have provided a web-based power calculation tool for determining the minimum number of twin pairs required to detect A and C in an ACE model of given parameters. Unfortunately, all published twin studies of attachment seem to be underpowered for detecting heritability (A), but most are powerful enough for detecting large environmental contributions.

In addition, behaviour-genetic analyses usually employ main effects models dividing up the total phenotypic variance into additive (or dominant) genetic, shared and non-shared environmental components. They normally do not separate gene-environment interactions (genetic sensitivity to environments) and gene-environment correlations (arising from experiences correlated with genetic propensities). Finally, quantitative genetic modelling does not go beyond the extent to which genetic factors influence behavioural traits, that is they are not informative about specific genes or specific environmental factors affecting the behaviours in question.

Molecular genetic studies of attachment

The development of molecular genetic methodology started to shift the focus of study towards locating and identifying specific genes underlying genetic effects evident in twin and adoptive studies, even before sequencing the human genome was completed. The widely used strategy of allele association investigates correlations of gene variants with phenotypes, i.e., it probes if individuals with a trait of interest carry a specific gene variant more frequently than individuals in a control group (case-control studies). The method of allele association has been quickly and productively applied in the area of multi-factorial mental illnesses for which genetic components have long been demonstrated by large heritability estimates. Genes regulating the cerebral levels of important neurotransmitters (dopamine, serotonin, GABA, etc.) or signal transmission efficiency (neurotransmitter receptors and genes) have been targeted in association studies of major psychiatric disorders such as schizophrenia, bipolar disorders, attention deficit/hyperactivity disorder (ADHD), and autism [84], as well as of personality traits [85].

The first investigation of the specific genetic background of attachment behaviour showed an association between D4 dopamine receptor (DRD4) gene polymorphism and infants' attachment behaviour [86]. The level of dopamine and the density of dopamine receptors in the PFC increase between 6-12 months of infant life, when many of these functions go through intensive development [87] and when the development of first attachment relationships typically takes place. The highly polymorphic DRD4 gene has a number of frequent functional variants in the populations [88]. One of these is a variable number 48 base pair tandem repeat (48 bp VNTR) in the coding region of the gene [89]. Receptor molecules coded by the 7-repeat allele have been found to have a lower potency for dopamine-mediated coupling to adenylate cyclase than receptors encoded by the other frequent, 2- or 4-repeat forms [90], and more recent results suggest an additional effect of the 48 bp VNTR on gene expression [91, 92]. Allele association studies have linked the 48 bp repeat polymorphism of the DRD4 gene with normal variations of neonatal, infant, and adult temperament [85], but also with clinical hyperactivity (ADHD) [93, 94].

In the longitudinal Budapest Infant-Parent Study (BIPS), a relative over-representation of the 7-repeat variant of the 48 bp VNTR polymorphism was found in the group of infants displaying insecure-disorganised attachment behaviour with their mothers in the Strange Situation [86]. The estimated relative risk for disorganised attachment among children carrying the 7-repeat allele was four-fold, with the frequency of the 7-repeat allele being 67% in disorganised infants as opposed to 20% in securely attached infants [95], and with 50% frequencies in the insecure-avoidant and resistant groups. Subsequently, we also reported an interaction between the structural 48 bp repeat polymorphism and the -521 C/T promoter polymorphism in the same group of infants: the association between disorganised attachment and the 7-repeat allele was enhanced by the presence of the -521T allele [96]. While Dutch researchers have failed to replicate this association [97, 98], parental genetic data and family-based analyses in the Hungarian sample showed a highly significant non-transmission of the 7-repeat allele (and the -521T

7-repeat haplotype) to securely attached infants, as well as a trend for preferential transmission to disorganised infants [95]. The preferential non-transmission of the 7-repeat allele to securely attached infants suggested that not carrying this allele might, in fact, have a protective effect, favouring the development of secure early attachment in this sample.

Gene-environment interaction effects on infant attachment

Specific genetic effects on phenotypes may be conditional on specific environments and thus be undetected in other environments. Similarly, exposure to specific environments are often influenced by choices that depend on the individuals' genetic make-up. Such interplay between genes and environment may channel individual development into different trajectories early in life and affect the long-term development of mental health and disorder. A shift from association studies targeting genetic main effects towards investigating the interaction of genetic and environmental influences is reflected by the accumulating evidence for such processes being reported in recent literature [99]. Important reports are emerging, describing the moderation of behavioural responses to early rearing conditions by specific genotypes. For example, in the Dunedin study, Caspi and colleagues found that links between childhood maltreatment and later psychological maladjustment were moderated by genetic factors [100]. The functional polymorphism of the monoamine oxidase A (MAOA) gene affecting enzyme activity moderated the relation between early maltreatment and later antisocial behaviour in males. In the same population, the 5-HTTLPR regulatory polymorphism of the serotonin transporter gene was shown to moderate the effect of early maltreatment on adult depression in both sexes [101]. Both findings have since been replicated [102–104]. Importantly, in these studies, the genetic factors had no main effects on the outcome and the genetic influence was detected only when the environmental measure of maltreatment was included in the analyses.

Our own study of the genetic effects on disorganised attachment has been extended by investigating the interplay between DRD4 gene polymorphism and maternal behaviour [54]. In order to increase the range of environmental and behavioural measures, we have done this in collaboration with the Boston-based US research group led by Dr. Karlen Lyons-Ruth, an expert on disorganised attachment in infants and atypical behaviour of mothers. Demographic risk, DRD4 7-repeat genotype, levels of atypical maternal behaviour and infant disorganised attachment were combined across the middle income, low risk Hungarian and the low income, high social risk US samples (N = 96 and 42, respectively). We found that, in the combined sample, disorganised attachment was related to both cumulative demographic risk and maternal atypical behaviour, but the main effect of infant 7-repeat genotype on disorganised attachment was no longer significant. However, the relation between maternal atypical behaviour and infant disorganisation was moderated by infant DRD4 genotype. The relationship was strong in the group of infants lacking the 7-repeat variant as expected, mothers showing a low level of anomalous behaviour had infants displaying a low level of disorganisation, and conversely, infants of highly atypical mothers showed a high level of disorganised attachment behaviour. In contrast, the level of disorganisation in infants who carried the 7-repeat allele was at an intermediate level and unrelated to the degree of maternal atypical behaviour. Infants carrying the 7-repeat allele thus seemed to be less sensitive to maternal behaviour. At the same time, the previously reported main effect of the 7-repeat genotype on attachment disorganisation was restricted to the group with mothers low on atypical behaviour. This pattern of results was also present in the separate US and the Hungarian samples. We hypothesised that functional variations in the DRD4 gene expressed preferentially in brain regions of the reward circuit might modulate sensitivity to maternal stimuli, which in turn might result in infants' differential sensitivity to aspects of care-giving behaviour.

Results of this first molecular genetic study of infant attachment [86, 96] seem to have transformed the attachment field by increasing interest in studying genetic and gene-environment interaction effects. A number of research groups have genotyped participants of previous and ongoing attachment studies, or have begun including genetic markers in the design of new studies. A further investigation of the DRD4 48 bp VNTR showed that infant genotype moderated the previously reported intergenerational transmission of mothers' unresolved trauma to disorganised attachment: the link was significant for infants carrying the 7-repeat allele only [105].

Studies have been extended by investigating associations with other candidate genes. The polymorphic serotonin transporter gene, coding for one of the important regulator of the synaptic level of the neurotransmitter serotonin, has been linked previously to anxiety-related traits [106] and affective disorders [107]. The 5-HTTLPR repeat polymorphism in the promoter region of the gene affects gene expression, that is the number of available serotonin transporter molecules. The 'short' 5-HTTLPR allele has been associated with low serotonin metabolism and behavioural problems in infant and juvenile Rhesus monkeys reared with peers only, but not in monkeys who were reared with their mothers and peers during infancy. In contrast, monkeys carrying only the 'long' allele showed normal metabolism and behavioural functioning, regardless of their early rearing history [108]. A similar protective effect of the homozygous 'long/long' (l/l) genotype or else a buffering effect of maternal behaviour was found in human studies of infant attachment. For infants with a 'short' allele, variation in mothers' responsiveness was significantly associated with attachment security [109] or attachment disorganisation [110, 111]. For infants homozygous for the 'long' allele (l/l), there was no association between maternal responsiveness and attachment security or disorganisation.

The gene-environment interaction effects on attachment reported in the above-cited publications are consistent with Belsky's differential susceptibility hypothesis [112], i.e., children's susceptibility to care-giving experience seems to be moderated by genetic factors. Unfortunately, all these attachment studies, as usual, involved relatively small sample sizes. In the light of the latest meta-analyses questioning the 5-HTTLPR main effects [113] and 5-HTTLPR by stressful environment interaction effects [114], these initial investigations of genetic and gene-environment interaction effects on attachment need to be confirmed by larger studies, possibly including multiple polymorphisms of multiple candidate genes.


Results and Discussion

An Excess of X-Derived Retrogenes in Eutherian and Marsupial Genomes

To trace the evolution of gene movements in mammals, we first screened for intronless retroposed gene copies (retrocopies) and their parental genes in three eutherian (“placental” mammal) genomes and one metatherian (“marsupial”) genome (opossum), using a refinement of our previously described procedure [2] (Materials and Methods). This analysis identified several thousand retrocopies in each of the therian genomes analyzed (Table 1). Thus, the process of retroposition has significantly shaped, not only the genomic landscape of eutherians [1,2], but also that of its sister lineage, the marsupials.

Retroposition in Therian Genomes

We then extracted two subsets from these retrocopy data for each species (see Materials and Methods for details): One was enriched for functional retrocopies (retrogenes Tables 1 and S1–S4), whereas the other contained retropseudogenes with open reading frame disruptions (premature stop codons and frameshifts) that likely preclude gene function (Table 1).

The analysis of chromosomal locations of parental genes revealed that X-linked genes of all genomes analyzed have spawned a large excess of functional retrogenes compared to autosomal genes, whereas no such bias is observed for parental genes that gave rise to retropseudogenes (Table 1). Thus, preferential fixation of functional X-derived genes by natural selection occurred, not only in eutherians [1,2], but also in metatherians. The latter is consistent with a recent study that showed that MSCI occurs in marsupials [12].

Autosomal Retrogenes Compensate for the Transcriptional Silencing of Their X-Linked Parental Genes

Before examining the evolutionary history of X movement patterns in more detail, we sought to obtain further evidence for the hypothesis that MSCI is the driving force for the preferential copying of genes from the X to autosomes, which was so far based on the analysis of individual genes (see Introduction). To this end, we analyzed expression patterns of retrogenes and their parental genes using genome-wide murine expression data [13] from testicular germ-cell populations, total testis, ovary, and 14 somatic tissues (Figure 1A Materials and Methods).

Log2 expression signals are represented according to the plotted scale. The two heat maps show signal values for both parental genes (left side) and their respective retrogenes (right side), or only for one of the two, when data are not available for the other. Upper and lower parts of the panels contain pairs with parental genes located on the X chromosome and the autosomes, respectively. Line numbers in the middle of the heat maps correspond to mouse retrogene identifiers in Tables S2 and S5.

We find that all parental genes are broadly expressed (median: 16, mean: ∼14.2 tissues), in significantly more tissues than other genes in the genome (median: 15, mean: ∼11 tissues, p < 10 −11 , Mann-Whitney U test Figure 1A), which substantiates previous notions that retrogenes stem from housekeeping genes with important functions in all or most tissues [2,7]. In contrast, the majority of X-derived retrogenes (12 of 17, ∼71%) are specifically expressed in testes (Figure 1A and Table S5). X-derived retrogenes show a striking excess of testis-specific cases compared to their parental genes (0 of 21 specifically expressed in testes) or other genes in the genome (790 of 14,991, 5.3% p < 10 −17 , Fisher exact test). We note that similar patterns have been described in Drosophila [14], a genus in which the out-of-X movement of genes was originally observed [15]. X-derived retrogenes in our data are also significantly more frequently expressed (specifically or nonspecifically) in testis (17 of 17, or 100%, with testis expression) compared to other, autosome-derived retrogenes (41 of 53, or ∼77%, with testis expression two-tailed p < 0.05, Fisher exact test 26 of 53, or 49%, are testis-specific). This points to a selective enrichment of testis functions among X-derived retrogenes during evolution, although retrogenes generally seem to be frequently expressed in testis, consistent with previous studies [2,3].

In order to functionally compensate for their parental genes in testes (Figure 1A), expression of X-derived retrogenes would be specifically required in testicular meiotic germ cells (spermatocytes), where their parental genes are silenced, but not in premeiotic spermatogonia (Figure 1B). Our expression analysis of premeiotic, meiotic, and postmeiotic cells revealed a striking pattern (Figure 1B and Table S5), consistent with a compensation function of retrogenes during but—surprisingly—also after meiosis (see [6] for recent evidence of active postmeiotic silencing of the X). In spermatogonia, X-linked parental genes show high and their retrogene copies low expression activity. Conversely, X-derived retrogenes are highly expressed in spermatocytes and postmeiotic spermatids, while their parental genes are silenced.

The overall propensity of retrogenes—including retrogenes with autosomal progenitors (Figure 1B)—to be expressed in spermatocytes/spermatids is probably due to the “hypertranscription” state of autosomal chromatin in these cell types ([3] and references therein). This likely facilitated the initial transcription of retrocopies after their emergence, allowing them to obtain functions in the late stages of spermatogenesis.

However, X-derived retrogenes are more frequently expressed in spermatocytes than retrogenes with autosomal parental genes (16 of 17, or 94%, X-derived vs. 39 of 53, or 74%, autosome-derived one-tailed p < 0.1, Fisher exact test), and at higher levels (median log2-transformed expression signal: ∼10.9 vs. 8.8, one-tailed p < 0.05, Mann-Whitney U test). A similar pattern is observed in postmeiotic spermatids (100% vs. 75% expressed, two-tailed p < 0.05 median signal: ∼11.0 vs. ∼10.5, one-tailed p = 0.15). In addition, based on expression cluster analyses [13] (Materials and Methods), we find that a significant excess of X-derived retrogenes show transcriptional induction in meiosis when compared to retrogenes that stem from autosomes (10 of 17, or ∼59%, vs. 16 of 53, or ∼30% two-tailed p < 0.05, Fisher exact test).

Our expression analyses substantiate the hypothesis that retrogenes that stem from the X have been fixed during evolution and shaped by natural selection to compensate for parental (housekeeping) gene silencing during (and after) MSCI. Thus, sexual antagonism (i.e., evolutionary conflict between males and females), which was previously considered as an alternative driving force for the fixation of X-derived retrogenes [1,16], likely played less significant roles for the selectively driven export of X-linked genes in mammals (at least for those that are specifically expressed during/after meiosis). In contrast to the mammalian pattern, X chromosome inactivation during spermatogenesis does not seem to be a major contributor to the out-of-X movement of genes in Drosophila [17]. Rather, it appears that the increased residency time of the X chromosomes in females accounts for the observed pattern in this genus [17]. Thus, interestingly, the predominant selective forces associated with the export of X-linked genes appear to differ between fruitflies (sexually antagonistic selection) and mammals (MSCI).

Gene Movements Reveal the Evolutionary Onset of Meiotic Sex Chromosome Inactivation

To date the evolutionary onset of the out-of-X movement of genes in mammals, we screened for the presence/absence of human retrogenes in genomes representing the three major mammalian lineages (see Materials and Methods for details). In addition to three eutherian and one marsupial genome (opossum), this analysis included a genome (platypus) of the most basal mammalian lineage, the egg-laying monotremes (Figure 2).


Conclusion and Discussion

This analysis offers three main findings. First, parents provide more cognitive stimulation to the twin that shows higher levels of early ability. This finding is in line with results by some previous studies reporting reinforcing parental behavior (Aizer and Cunha 2012 Behrman et al. 1994 Datar et al. 2010 Frijsters et al. 2013 Rosales-Rueda 2014 Yi et al. 2015). It is also consistent with the intrafamily resource allocation model introduced by Becker and Tomes (1976), suggesting that parents reinforce human capital differences between their children but may compensate with direct nonhuman capital transfers later in life. Although this empirical finding is clear and consistent across alternative measures of cognitive ability, the interpretation is not conclusive. Parents may not be guided by the desire to reinforce ability differences but rather may find it easier, more pleasant, or more rewarding to interact with the child with higher ability. Further research including parental reports should help to clarify this point.

The focus of our analysis is on the socioeconomic heterogeneity in parental responses. The second finding speaks to this issue. We show that high-SES parents provide more cognitive stimulation to the higher-ability twin, whereas low-SES parents do not appear to respond to differences in their children’s early skills by either compensating or reinforcing behavior. This finding is consistent across all measures of parental socioeconomic advantage (parental education, household income, and SES) and across two measures of children’s early ability. This finding highlights the stratification of parental childrearing orientations. High-SES parents may be more likely to notice and scrutinize ability differences between their children (Lareau 2011). In any case, these results depart from previous studies based on sibling fixed-effects models, which report that lower-class parents reinforce birth weight differences by investing more in the child with higher birth weight whereas upper-class parents compensate by investing more in the child with lower birth weight (Hsin 2012 Restrepo 2016). In combination with prior research, our findings suggest that parents react differently to different types of child’s endowments and that we cannot generalize results based on birth weight (Almond and Mazumder 2013).

The third finding from our study is that the additional cognitive stimulation that advantaged parents offer to their higher-ability twin does not widen differences in children’s cognitive (or socioemotional) performance at age 4 or 5. The null association between parental cognitive stimulation and children’s later cognitive outcomes adds to a nascent body of research examining the causal effect of parental involvement on child outcomes. Although this finding is not conclusive because the null effect may be partly related to measurement error, it provides suggestive evidence that parental cognitive stimulation early in life does not widen inequality within families. Much survey research has studied the association between different dimensions of parental involvement and children’s cognitive, behavioral, and emotional outcomes (El Nokali et al. 2010 Harris et al. 1998 Kaushal et al. 2011 Milkie et al. 2015 Wang et al. 2014 Yeung et al. 2002), but virtually all these studies were cross-sectional and could not rule out confounding emerging from unobserved family- or pregnancy-specific factors. More conclusive evidence on the effects of cognitive stimulation and parental involvement during early childhood has recently emerged from randomized interventions that demonstrated positive short- and long-term effects (Attanasio et al. 2014 Gertler et al. 2014). These interventions are usually targeted to disadvantaged populations in developing nations, providing a blueprint for further analysis in the United States and other contexts that would add to evidence using twin fixed-effects models or other causal inference strategies to better account for unobserved heterogeneity.

Recent scholarship using child fixed-effects models to isolate the effect of children’s time spent with mothers on their cognitive outcomes finds that educationally oriented time (but not total parental time) that children spend with their mother leads to higher test scores (Hsin and Felfe 2014). Our findings of a null effect of parental cognitive stimulation depart from the positive influence detected by Hsin and Felfe (2014). Different findings could be explained by differences in methodological approaches (twin fixed effects or child fixed effects), measures of parental involvement (parental cognitive stimulation or time spent with children), and ages at which the outcome is measured (early or late childhood). Diverging results highlight the need to expand this nascent body of research in order to understand when and which types of parental involvement benefit children’s developmental outcomes, and whether these effects vary by SES.

Our findings suggest that a more sophisticated understanding of parental involvement may be needed, moving beyond an indistinct notion of “parental responses” given that parents may react differently to different types of child endowments. Research would also benefit from including diverse dimensions of involvement, including time spent with children, quality of time, cognitive stimulation, and investment in goods and educational materials.

We motivated this study by suggesting that parental responses may matter for the intergenerational transmission of advantage by inducing within-family inequality. What do our results imply for parental involvement as an underlying mechanism of intergenerational persistence? Because we do not find evidence that greater parental support to the higher-ability twin affects later cognitive outcomes, parental behavior—even if reinforcing in its allocation—is not found to be reinforcing in its consequences in this analysis. However, the main finding of this article remains that parental responses vary by SES and that advantaged families engage in reinforcing responses to differences in their children’s cognitive skills, which in turn may affect children’s educational achievement and attainment later in life. We hope that other studies will address this possibility, furthering a more comprehensive understanding of the stratification of parental responses and their consequences throughout the early life course.


Results

A systematic framework for inferring gene functions from high-dimensional phenotypic data

KCML aims to utilise existing biological knowledge, as captured by GO, to automatically identify gene perturbation phenotypes in HT-GPS datasets and map these phenotypes to potential biological functions. Central to our approach are GO term classifiers that effectively link gene annotations, which are not specific to a cell type or biological context, to the rich contextual information provided in HT-GPS datasets. Critically, each term classifier identifies the phenotypic signature associated with a given gene annotation, and it can therefore be used to study how perturbation of biological functions at lower-scale contributes to higher-scale phenotypes.

Here, a weakly supervised learning approach is used to train each term classifier. Binary support vector machine (SVM) classifiers prove to be well suited to this problem (Materials and Methods). Each GO term classifier learns to discriminate between perturbation phenotypic profiles of genes annotated to that term (positive class) and a random set of remaining genes (negative class) (Fig 1A). To select features that are relevant to a given biological function, we initially train the GO term classifier using one feature and iteratively add features to the model if they improve the performance of the classification based on the F-score metric (i.e., forward feature selection see Materials and Methods). Only those classifiers that exceed a certain performance threshold on unseen data are used for prediction (Materials and Methods). For example, a classifier of the GO term “cell cycle” will select features that are discriminative of the perturbation effect of the annotated cell cycle genes and define a decision boundary separating these effects from a random set of negative genes (Fig 1A). If such a classifier can successfully predict a held-out sample of annotated cell cycle genes, then it is used to predict other potential cell cycle genes based on phenotypic similarity. These predictions can be ranked based on the distance to the SVM decision boundary, which indicates the strength of the phenotype. Predictions from different GO term classifiers are then combined to yield a data-driven multi-ranking of gene functions. One of the main advantages of this approach is the effective augmentation of GO annotations in a context-dependent manner.

Figure 1. KCML workflow for inferring gene functions from different large-scale genetic datasets

  • A. KCML workflow trains multiple classifiers to identify phenotypes and genes associated with different gene ontology terms. Feat: feature.
  • B. Categories and numbers of included GO terms.
  • C. The distribution of genes per term and vice versa based on GO annotations.
  • D. Tested datasets.
  • E, F. (E) Overview of single-cell-resolved image-based features measuring morphology, microenvironment and infection (F) and the generated statistics based on a population of single cells when applicable (n > 625 and Table EV2). Examples on the various measurement types are shown on the right and include multiple quantiles, mean, range, IQR (interquartile range), kurtosis, skewness, bimodality and KS distance.

We train GO term classifiers from the three GO ontologies: biological process, molecular function and cellular component (Fig 1B, Materials and Methods and Table EV1). As GO terms can vary in specificity and number of annotated genes depending on their position in the GO tree, we only included terms with a sufficient number of positive examples (100–500 annotated genes). On average, each gene has 30 annotations (Fig 1C) with 2,152 genes having more than 50 annotations and 2,908 genes having no annotations based on the selected terms.

KCML generalisability

To illustrate the generalisability of KCML to various genetic perturbation screens, we applied it to three datasets measuring different types of phenotypic data and utilising different experimental technologies: (i) pooled genome-wide CRISPR/Cas9 screens measuring cell viability in 60 cancer cell lines (cell population-level phenotype) (Rauscher et al, 2018 ) (ii) a large-scale siRNA screen measuring changes in the expression of 3,287 genes in MCF7 breast cancer cells (transcriptome-level phenotype) (Duan et al, 2014 ) and (iii) an arrayed image-based genome-wide siRNA screen measuring changes in 168 single-cell features that quantify morphology, microenvironment and infection in HCT116 colorectal cancer cells based on stains of DAPI and the rotavirus-expressed viral protein 6 (VP6, single-cell and population-level phenotypes) (Green & Pelkmans, 2016 ) (Fig 1D and E). The latter dataset is composed of single-cell measurements, in which each perturbed population has 6,040 cells on average (Fig EV1A). To capture the heterogeneity in cellular states and collective cellular behaviour as readouts, we aggregated single-cell features into 1,719 features per gene perturbation profile. Computed statistical measures include standard deviation (SD), various quantiles, skewness, kurtosis, bimodality coefficient, Kolmogorov–Smirnov (KS) (Altschuler & Wu, 2010 ) and rank-sum statistics (Fig 1F, Materials and Methods and Table EV2). Furthermore, we corrected all features to detect cellular phenotypes independent of cell number (Fig EV1A–D and Materials and Methods).

Figure EV1. Single-cell image-based data allows dissecting perturbation effects that are independent of cell number and complements other types of phenotypic data

  • A. Estimated density of the number of cells across 18,033 gene knockdowns.
  • B, C. Scatter plots of TCN (total cell number) versus Nuclei area mean (B) and II mitotic (infection index of mitotic cells) before and after correction for dependency on TCN (C). R 2 indicates Pearson correlation coefficient. Colour indicates number of cells (0–2,000, … 18,000–20,000).
  • D. Box plots of the correlation between SVM confidence scores and cell number before and after TCN correction. Box plots elements: centre line, median box limits, 25 th and 75 th percentiles whiskers, ± 2.7 standard deviation. Points: outliers (n = 145).
  • E. Individual AUROC curves for the different GO term classifiers for the three datasets.
  • F. GO terms where single-cell-resolved image-based readouts outperform expression readouts based on overall recall.
  • G. GO terms that are only classifiable based on the image-based dataset.
  • H. Precision–recall curves for the three datasets where the shaded area indicates the 95% confidence interval (left). The same is shown on the right but when considering first neighbours of annotated genes for a given GO term in the protein–protein interaction network as true-positives.

Performance and validation of KCML

For each of the datasets, we selected the most confident GO term classifiers based on the recall (true-positive rate) and false-positive rate such that it is significantly better than classifying a random set of genes (Fig 2A and Materials and Methods). The average area under the ROC curve (AUROC), which reflects specificity versus sensitivity, for the three datasets is 75.44% (Figs 2A and EV1E). The HT-GPS dataset using gene expression as a readout had the highest recall and AUROC, as well as the number of classifiable terms—i.e., classifiers scoring above our selection threshold (Fig 2A and B). This is expected since this dataset measures, for each perturbation, changes in the expression of 3,287 genes, whose variation is representative for most of the transcriptome (Duan et al, 2014 ). Interestingly, viability measured in 60 cancer cell lines from 12 tissue types can also be informative of gene functions as the number of classifiable terms is comparable with the multivariate image-based screen (Fig 2A and B).

Figure 2. KCML performs best in gene function inference and predicts many novel gene functions

  • A. AUROC and recall of test samples of classifiable terms across the three tested datasets (n > 141).
  • B. The number of classifiable terms in the three datasets.
  • C, D. Comparison of KCML against classifiers that were trained to classify random sets of 200 genes (P < 1.8e-17) (C), or clustering of genes using k-means, SOM (self-organising maps) or enrichment for pairwise correlations (P < 1e-150) (D) (n > 100).
  • E. Validation annotations are defined based on annotations accumulated between Mar 2018 and July 2019.
  • F. Percentage of new GO annotations that was predicted by KCML correctly.
  • G. Correlation between the performance of GO term classifiers on training annotations (annotations2018) versus validation annotations (annotations2019-2018).
  • H. Predictions of many GO terms are significantly enriched for first neighbour interactions based on STRING, physical (based on experimental evidence in STRING) and Pathway Commons databases (hypergeometric test). Error bars indicate 1 standard deviation (n = 145, 141, 688 from left to right).

Multivariate single-cell-resolved image-based readouts outperformed gene expression for some GO terms, while 33 terms were only classifiable based on image-based single-cell features (Fig EV1F and G). Many of those are involved in phosphorylation, ubiquitination or membrane transport. This might be due to the fact that bulk gene expression does not necessarily capture phenotypic effects caused by post-transcriptional events, or measure changes at the single-cell level. For example, exopeptidase activity and voltage-gated channels had a higher overall recall based on image-based single-cell features compared to the gene expression dataset (Fig EV1F). These results show that different experimental techniques for probing biological systems complement each other and provide different functional information.

The GO term classifiers obtained using KCML captured biologically relevant signatures, as they performed significantly better than classifiers based on random sets of genes (Fig 2C, P < 1.8443e-17 and Materials and Methods). Furthermore, KCML outperformed commonly used analysis pipelines for HT-GPS datasets based on dimensionality reduction and unsupervised clustering. Clustering of gene profiles using k-means and self-organising maps while varying the number of clusters resulted in AUROC of 50%, which indicates random classification (Fig 2D and Materials and Methods). Similar results were obtained even when we considered pairwise correlation between genes in a given GO term (Fig 2D and Materials and Methods). These results confirm the power of KCML in identifying functionally relevant phenotypes and highlight the need for such a systematic approach, as existing analysis methods do not provide a clear functional interpretation of quantitative phenotypes.

To obtain unbiased validation of the KCML predictions, which were based on annotations downloaded from UniProt in March 2018 (annotations2018), we downloaded annotations accumulated in UniProt in July 2019 (annotations2019) and used the new annotations between March 2018 and July 2019 (annotations2019-annotations2018) as a validation set (11,383 annotations) (Fig 2E). Since these new annotations were never seen by KCML, and were based on diverse resources, they are not likely biased to a particular dataset. We found that KCML can significantly predict many of the newly reported annotations with comparable performance for the different datasets (Fisher's exact test right-tail P < 0.025e-11 and Fig 2F). Furthermore, the performance of our models on annotations2018 correlates with the performance on the new UniProt annotations (Fig 2G). Therefore, GO term classifier performance can be considered for selecting the more likely hypotheses.

An example of a gene for which new UniProt annotations are consistent with KCML predictions is URB1 a homolog of the essential Npa1p yeast gene. URB1 has been characterised to play a role in early steps of 60S ribosomal subunit biogenesis via associating with RNP proteins, and it localises to the fibrillar centre of the nucleolus, where ribosome biogenesis takes place (Uhlén et al, 2015 Farley-Barnes et al, 2018 ). In August 2018, UniProt had assigned URB1 a new function: non-coding RNA processing based on a phylogeny analysis. KCML predicted that URB1 is involved in non-coding RNA processing from each of the three datasets analysed, which were generated by different laboratories and different gene perturbation technologies. URB1 interacts with 26 ncRNA-processing proteins based on STRING and Pathway Commons databases, which further supports this prediction.

Other examples of new UniProt annotations that are consistent with KCML predictions include the following: (i) All members of CD1 glycoprotein family (CD1A, CD1B, CD1C, CD1D and CD1E) were correctly predicted to be associated with “regulation of the adaptive immune response” based on the siRNA screen with transcriptome readout and (ii) the mini-chromosome maintenance (MCM) complex proteins (MCM2-7), which are known to be involved in DNA replication, were correctly predicted to also be involved in double-strand break repair, based on both the CRISPR/Cas9 cell population viability screens and the siRNA screen with transcriptome readout. These examples illustrate the power and generalisability of KCML in predicting new gene functions based on gene perturbation screens which can be useful for generating data-driven hypotheses.

As gene ontology annotations are noisy and incomplete, KCML can generate a high number of potential hypotheses. We sought to determine whether these predictions can be due to the propagation of perturbation effects to neighbouring genes in the protein–protein interaction network. Especially, it is expected that interacting genes are more likely to be functionally related and their depletion to result in a similar phenotype (Evans et al, 2013 ). We found that a large proportion of predicted genes for a given GO term do interact with the genes that are previously annotated to that term, based on STRING and Pathway Commons databases (Fig 2H). This enrichment is significant for many of the terms. Moreover, accounting for protein–protein interactions results in a large improvement in precision–recall curves (Fig EV1H). This provides further validation of KCML predictions, where neighbouring genes in the interaction network are more likely to perform similar functions.

Using KCML to identify GO terms represented in the image-based dataset and their relationships

To demonstrate the use of KCML, we focus on the further analysis of the image-based perturbation dataset, as it is the only dataset that provides spatially resolved single-cell measurements, which are particularly challenging to analyse and interpret (Collins, 2009 ). We trained KCML using a subset of features to determine the GO terms that can be learned from different cellular markers (morphology and microenvironment features based on DAPI versus infection features based on VP6 staining) (Materials and Methods). Shape features are predictive of 61 GO terms while infection measurements are predictive of 38 terms, with 8 terms shared (Figs 3A–C and EV2A). Combining both feature sets result in an additional 54 classifiable terms (Fig 3A). These results illustrate that multivariate imaging data can be predictive of many gene functions even when only one or two cellular stains are used.

Figure 3. Analysis of functional information using KCML based on image-based dataset

  1. The number of classifiable terms when using morphology versus infection features.
  2. AUROC and recall of test samples for classifiable terms when using different subsets of features (n > 38). Box plots elements: centre line, median box limits, 25 th and 75 th percentiles whiskers, ± 2.7 standard deviation. Points: outliers.
  3. The improvement in performance when all the features are used for terms that are classifiable when only shape features are used (n = 61), only infection features are used (n = 38), or either shape or infection features are used (n = 8). Only a slight improvement in classification performance is achieved by combining feature subsets when the signatures based on a subset of features are already strong. Error bars represent mean + SD.
  4. Network representation of classifiable terms where edges indicate the overlap in the predicted gene lists between GO terms.

Figure EV2. The predictions of GO term classifiers have a moderate overlap and select a diverse range of features

  • A. The recall on test data for terms that are classifiable based on both infection and shape features.
  • B, C. Clustering (B) and distribution (C) of Jaccard index values based on the overlap between predictions of GO term classifiers. Most terms have a moderate overlap suggesting that different phenotypes are discovered.
  • D. The number of features in different categories that are selected by the respective GO term classifier (scaled) when the classifier is trained using (i) only shape features or (ii) only infection features. Blue indicates the number of features with a higher average than control, while red indicates the number of features with a lower average than control.
  • E. Feature categories based on feature type (e.g., morphology, cell context, DAPI intensity), measurement type (e.g., summary, spread, distribution shape). The topmost frequently selected features by different GO term classifiers are shown as examples if applicable. Hue of orange circles indicates feature importance based on the number of classifiable terms where the corresponding feature was selected.

To gain insight into the functions that have been learned by KCML, we generated a network of classifiable GO terms based on the overlap in their predicted gene lists (Figs 3D and EV2B and C). We observe a strong cluster of membrane transport-related terms including potassium, calcium, sodium and metal ion channels (Fig 3D). Most of these terms can be classified using morphology and microenvironment features alone and their phenotypic profiles cluster together (Figs 4A and EV3). Interestingly, based on our data, ion channel terms (molecular-level) are linked to multicellular organismal signalling function (tissue level). Another cluster included many phosphorylation- and ubiquitination-related terms, many of which are classifiable based on infection features (Fig 3D). Thus, KCML can classify biological functions that act at different scales, from the molecular to the tissue level.

Figure 4. Association between phenotypic changes in single-cell-based dataset and GO terms

  1. Hierarchical clustering of the average fold change of predicted positive versus negative samples for each term using all features. Hierarchical clustering is based on ward linkage and Euclidean distance. Regions outlined by cyan rectangles are shown in Figs EV3 and EV4A.
  2. The number of features in different categories that are selected by the respective GO term (scaled). Blue indicates the number of features with a higher average than control, while red indicates the number of features with a lower average than control. * indicates the GO terms that are classifiable by either shape or infection features.
  3. Example cell images following knockdown (k/d) of known or predicted MSD genes versus control (scrambled). (ii) zoom-in image of the region highlighted in (i). Blue: DAPI, and green: VP6 antibody. Scale bars = 65 μm.

Figure EV3. Morphology and microenvironment features are predictive of many GO terms involved in membrane transport

  • A, B. Examples of significantly changed features for terms in Fig 4A (C1) which include many membrane transport terms. TCN: total cell number.

KCML automatically maps GO terms to functional phenotypes

To understand the phenotypic changes associated with different GO terms, we categorised the phenotypic features based on feature type (cell context, morphology, DAPI intensity and texture, VP6 intensity and texture, state, etc.) as well as measurement type (summary statistics, spread statistics, distribution shape or distribution distance—the distance between perturbed and control distributions) (Fig EV2E and Table EV2). We then counted and scaled the number of features selected by a term classifier in the different categories (Figs 4B and EV2D). Notably, the rank-sum statistic was among the most selected measure by different classifiers indicating its biological relevance and robustness (Fig EV2E).

Membrane transport-related terms are predicted based on the increased number of cells and total cell area as well as local cell density and cell distance to islet edge (Figs 4B and EV3). Surprisingly, the “ligand gated ion channel activity” and “voltage-gated ion channel activity” terms are the most accurately classified (Fig EV2A). Depletion of chloride transport genes also affected many texture and intensity measures of VP6, but most interestingly, it significantly decreased rotavirus infection (P < 4.9e-139, Figs 4B and EV3A). These results illustrate the functional relevance of single-cell aggregated measurements of cell morphology and context and suggest that chloride channels might be required for the spread or replication of rotavirus.

From image-based data, KCML predicted a number of genes involved in mesoderm development (MSD). This was based on the spread and summary values of cell context, DAPI intensities, as well as the distribution shape of cell morphology, which are indicative of changes in cell organisation (Figs 4B and EV4A). Indeed, depletion of previously annotated (e.g., SMAD3) and predicted (e.g., ITGAV, WNT8B and OR51B4) MSD genes results in small cell colonies compared to control (Fig 4C). This phenotype might indicate the inability of cells to spread or migrate. This hypothesis is supported by the high overlap of the MSD-predicted genes with GO terms such as migration (epithelial and ameboidal-like) and morphogenesis of branching structure (Figs 3D and EV4A). Thus, through a holistic analysis of multiple features, KCML reveals a role for MSD-associated genes in the spread and organisation of epithelial colorectal cancer cells into a uniform epithelial sheet.

Figure EV4. Microenvironment and heterogeneity measurements are predictive of Mesoderm Development GO term and their variation can specify lower-scale functions

  1. Example of significantly changed features for terms in Fig 4A (C10) that include MSD.
  2. Average values of features selected by MSD classifier for genes predicted to perform additional functions other than MSD.

KCML allows multi-scale analysis of high-dimensional data

Because KCML predicts GO terms associated with different biological scales, we asked whether it can subdivide higher-scale properties shared by a set of genes into gene subsets that carry out different aspects of the higher-scale property at lower scales. For example, previously annotated genes to a higher-scale GO term such as MSD are predicted by KCML to perform different functions at lower scales, such as positive regulation of epithelial cell migration and mammary gland development (Fig 5A). Interestingly, MSD genes that are predicted to play different functions at lower scales tend to occupy different regions in the MSD subphenotypic space, reflected by differences in a subset of MSD features (Figs 5B and EV4B). Thus, KCML allows specifying a common higher-scale phenotypic property emerging from the collective action of a set of genes into different lower-scale aspects of this property performed by subsets of these genes.

Figure 5. KCML allows pleiotropic analysis of high-dimensional phenotypic data

  1. Heatmap showing SVM-based ranks (z-scored) for known MSD genes against multiple functions. The functions on the y-axis are the top ten overlapping terms with MSD classifier prediction.
  2. Embedding of MSD subphenotypic space using t-SNE based on the selected features by MSD classifier where only MSD genes are considered (Materials and Methods). Colour indicates the respective SVM rank for each gene for the corresponding function.

In addition, perturbation of a biological function often affects only a subset of the measured features. For example, the mean and integrated intensity of DAPI is significantly lower in genes predicted to participate in cell cycle checkpoint (Fig EV5A). However, these genes do not cluster together when all measured phenotypic features are considered, but only when considering features selected by the cell cycle checkpoint classifier (Fig EV5B and C). Importantly, identification of subphenotypic effects associated with each gene allows determination of which of those subphenotypic effects are potential off-target effects. Specifically, predictions of genes that are targeted by the same siRNA seed (guide strand) are filtered out if the seed is significantly over-represented in a given GO term classifier (Fig EV5D–F and Materials and Methods). This emphasises the importance of searching in the subphenotypic space to deconvolve biological signals contained in high-dimensional data and obviate off-target effects.

Figure EV5. Deconvolution of sub-phenotypic sapces and off-target effects

  • A–C. Comparison between negative and positive genes predicted by the respective GO term classifier based on: two significantly changed features (A), subphenotypic t-SNE embedding (based on the selected features by the respective classifier) where red hues indicate SVM rank (B), or phenotypic t-SNE embedding (based on all measured features) (C).
  • D. Example of predicted functions for five genes and the detected off-target effects based on siRNA seed enrichment.
  • E, F. Distribution of the number of off-target predictions per GO term classifier (E) or per gene (F).

Validation of mesoderm development classifier predictions in the context of colorectal cancer

Since loss of healthy tissue organisation is one of the main characteristics in tumours (Hinck & Näthke, 2014 ), we sought to analyse MSD genes whose perturbation resulted in abnormal cell organisation between HCT116 cells in culture. As expected, MSD genes are significantly enriched for morphology-regulating pathways including tight junctions, focal adhesion and actin cytoskeleton (Figs 6A and EV6 and Table EV3). Among the predicted genes are 15 collagen, 12 integrin and many polarity genes, such as Par3 and Par6 (Table EV3). Moreover, many of the predicted MSD genes also participate in pathways that are often dysregulated in colorectal cancer including TGFβ, WNT and PI3K-AKT pathways (Muzny et al, 2012 ) (Fig 6A). Predicted genes include TGFβR2, PTEN and ERBB2 which are often mutated in cancer (Kuipers et al, 2015 ). TGFβ and WNT signalling are known to contribute to mesoderm development and their over-activation is associated with mesenchymal and stemness phenotypes in colorectal cancer, respectively (Hinck & Näthke, 2014 ). Collectively, these results illustrate how KCML allows constructing an integrative view of how modular gene programmes coordinate different signalling pathways to drive cellular phenotypes.

Figure 6. MSD predicated gene list is associated with key colorectal cancer pathways and patient outcome

  • A. KEGG pathways that are significantly represented in MSD genes (P < 0.05 based on right-tail Fisher's exact test). Previously annotated genes (known) to MSD are shown in red while others are predicted by KCML based on phenotypic similarity to known MSD genes.
  • B. Network depicting the interactions between MSD genes based on STRING and Pathway Commons.
  • C. TCGA colorectal cancer patients’ data projected into the first two principal components based on the expression of all genes (left), and expression of the top 300 predicted MSD genes (middle) and the expression of known mesoderm genes.
  • D. Spearman correlation coefficient between TGFβ or WNT signatures and the average of known MSD genes, average of predicted MSD genes (rank 1–300, rank 301–600 and rank 301–1,000) or average of MSD-associated olfactory receptors (P < 0.05). ORs: olfactory receptors.
  • E, F. Survival of colorectal cancer patients (months) against the expression of OR51B4 (E) and OR5K1 (F). Colour indicates tumour grade where grade 1 is well differentiated, grade 2 is moderately differentiated, and grade 3 or 4 is poorly differentiated.
  • G–I. Kaplan–Meier survival analysis of colorectal cancer patients based on Grade 3 ± tumours and the expression state of OR51B4 (G), OR5K1 (H) and olfactory receptor metagene (I), which is aggregated, based on the expression of many olfactory receptors (Materials and Methods and Fig EV7G).
  • J, K. Wald statistic value based on Cox proportional hazard regression analysis of OR5K1 predictivity of survival against tumour grade, presence of lymph nodes and metastasis. Detailed results for all tested variables and models significance are shown in Tables EV5–EV7. * indicates P < 0.05, and ** indicates P < 0.001.

Figure EV6. Predictions of MSD classifier are significantly enriched for cell adhesion and colorectal cancer pathways

As known MSD genes are highly implicated in colorectal cancer, we sought to determine the relevance of predicted MSD genes in colorectal cancer patients. We interrogated The Cancer Genome Atlas (TCGA) gene expression dataset (Muzny et al, 2012 ) of 577 colorectal cancer patients. Four consensus molecular subtypes (CMS) of colorectal cancer have been identified: CMS1 (microsatellite instability), CMS2 (WNT activation), CMS3 (metabolic) and CMS4 (mesenchymal) (Guinney et al, 2015 ). Strikingly, the top 300-predicted MSD genes recapitulate colorectal cancer molecular subtypes with comparable performance to using all genes or known mesoderm genes (Fig 6C, Materials and Methods). This is significantly higher than random sets of 300 genes (P ≤ 3.44e-25 and Materials and Methods). Therefore, MSD genes that alter epithelial cell organisation in HCT116 cells can stratify colorectal cancer patient's molecular subtypes.

Next, we sought to validate the relationship between TGFβ and WNT signalling, and expression values of predicted MSD genes in colorectal cancer. Average expression of WNT genes is used as a surrogate for WNT signalling while the TGFβ signature is based on genes reported by a previous study (Calon et al, 2012 ) (Materials and Methods). We found a significant correlation between MSD genes and TGFβ and WNT signatures in colorectal cancer patients (Fig 6D). The correlation between the predicted MSD genes and TGFβ/WNT signalling in colorectal cancer patients further supports their functional interaction.

Strikingly, 83 olfactory receptors were predicted by KCML to be involved in MSD, and their average expression correlates with WNT and TGFβ signatures (Fig 6B and D, Table EV4). These include OR51B4 and OR5K1 genes that are over-expressed in HCT116 cells when compared to 947 cancer cell lines (Barretina et al, 2012 ). We sought to investigate the correlation between olfactory receptors and colorectal cancer patient outcomes as their application in colorectal cancer therapeutics is beginning to emerge (Lee et al, 2019 ). OR51B4 and OR5K1 are mostly expressed in patients with tumour grade 3 or higher (Fig 6E and F). Their expression, albeit at a low level, is predictive of significantly worse patient outcome based on Kaplan–Meier log-rank test especially in grade 3 and 4 tumours (Figs 6G and H, and EV7A–E). Similar results are obtained when we aggregated the state of many MSD-associated olfactory receptors (Materials and Methods and Figs 6I and EV7F and G). This significant association was not observed using randomly selected genes (Materials and Methods). Importantly, expression of OR5K1 or the olfactory receptor metagene is predictive of survival, independent of other clinical variables including tumour grade, presence of lymph nodes and metastasis state, as well as the expression levels of neighbouring oncogenes (Fig 6J and K, Tables EV5–EV7 and Materials and Methods). The expression values of these genes are unlikely to be due to misalignment with other olfactory receptors as they generally share less than 95% sequence similarity with other olfactory receptors (Fig EV7H and Table EV8). These findings confirm the association between expression values of olfactory genes and the WNT and TGFβ pathways, which is consistent with our KCML prediction based on gene perturbation phenotypes.

Figure EV7. Most olfactory receptors have low expression in colorectal cancer patients but correlate with patient survival

  • A, B. Distributions of the expression of all genes (A) and MSD-associated olfactory receptors (B) in colorectal cancer patients based on TCGA.
  • C. Distribution of the number of patients where an MSD-associated olfactory receptor is expressed.
  • D–F. Kaplan–Meier survival analysis of colorectal cancer patients based on the expression state of OR51B4 (D), OR5K1 (E) and MSD-associated olfactory receptor metagene (F). OR: olfactory receptor.
  • G. Derivation of MSD-associated olfactory receptor metagene where the expression state of each receptor is added iteratively to the metagene and retained if scored significant based on Kaplan–Meier survival test (validated by leave-one-patient-out).
  • H. Sequence similarity between olfactory receptors in our metagene with other olfactory receptors based on the human olfactory data explorer (Olender et al, 2013 ).

How the Mind Really Works

How the Mind Works is a synthesis of cognitive science and evolutionary biology that aims to explain the human mind with three ideas:

1. Computation: thinking and feeling consist of information-processing in the brain

2. Specialization: the mind is not a single entity, but a complex system of parts designed to solve different problems

3. Evolution: as with the organs of the body, our complex mental faculties have biological functions ultimately related to survival and reproduction.

The book lays out criteria for attributing an evolutionary function to a trait, and applies them to many hypotheses using data from cognitive science, psychology, anthropology, and biology.

In their review, Jeremy Ahouse and Robert Berwick dismiss it entirely. "Don't believe a word of it," they say about what they take to be its key assumptions. "More is not always more . . . it is sometimes disastrously less." Despite their vehemence, it is not easy to see what the substantive disagreement is about. Some scientists disagree with 1., and assert that the mind is a direct product of the biochemistry of the brain, but Berwick cannot be among them: he himself takes a computational approach to language and mind.1 Others disagree with 2., and assert that there is a generic neural network learning algorithm which the brain uses for every mental process. But Berwick cannot be among them, either: he works in a Chomskyan framework in which language is treated as a specialized module, and presumably he does not believe that language is the brain's onlyspecialized module. As for 3., evolution, Berwick cannot be opposed in principle to examining the phylogenetic basis of mental faculties, since he himself has recently done just that for language.2

There remains an issue of which mental processes are functional adaptations, as opposed to by-products of adaptations or the result of chance. It is unlikely that Ahouse and Berwick would deny that one can ever say that some part of the mind is an adaptation the idea that stereo vision, fear, and sexual desire are adaptations to see in depth, avoid danger, and beget children are as indispensible and uncontroversial as the idea that the heart is an adaptation to pump blood or the kidneys are adaptations to filter it. There may be disagreements over whetherparticular mental faculties (such as probabilistic reasoning, grief, or humor) are adaptations, and if so, what they are for, but such arguments do not appear in Ahouse and Berwick's review.

The review becomes a bit less mysterious when we recognize that their accusations are taken, often wording and all, from Richard Lewontin. The review is the latest example of a conventional genre in modern intellectual life: the all-out attack by Lewontin or his collaborators (including Steven Rose, Stephen Jay Gould, and Philip Kitcher) on attempts to connect psychology with standard evolutionary biology. Since 1975, when Lewontin and others published their "Against Sociobiology" manifesto claiming that such attempts were politically reactionary and encouraged eugenics and Nazism, the attacks have recycled the same accusations and tactics: fuzzy scare words ("atomism," "reductionism," "determinism"), misreportings and doctored quotations, shameless straw-manning, empty name-calling ("vulgar," "pop"), political smears, personal innuendo, and most of all, the scientific snow job: seemingly damaging technical findings that general readers are unlikely to know about and hence unlikely to recognize as red herrings.3

How the Mind Works (HTMW) tried to preempt this kind of attack by patiently raising and refuting all the standard criticisms, both political and scientific. But Ahouse and Berwick (hereafter A&B) seem determined to discredit the book by any means necessary, and so rather than engaging its arguments, they have redoubled the use of shoddy tactics. As an example of their standards of fairness, here is an excerpt in which they try to show that I am ignorant of basic evolutionary biology:

Pinker asks, "Why don't women give virgin birth?" Certainly the correct answer will not make particular reference to humans. Mammals, including humans, just do not have this as a developmental option. Put otherwise, it would take more than just a shift in selection regimes to make humans start reproducing asexually. For these reasons it is hollow bluster to talk about the selective advantage of sex in humans if the traits we are discussing evolved and became established long before the human lineage branched.

This guy Pinker sure does sound naive! Unfortunately for the readers of Boston Review, the "hollow bluster" is a sentence fragment that A&B chopped out of context:

Why is there sex to begin with? . . . Biologically speaking, the costs are damnable indeed, so why do almost all complex organisms reproduce sexually? Why don't women give virgin birth to daughters who are clones of themselves instead of wasting half their pregnancies on sons who lack the machinery to make grandchildren and are nothing but sperm donors? Why do people and other organisms swap out half their genes for the genes of another member of the species, generating variety in their offspring for variety's sake? (pp. 461-62).

The discussion that follows also makes it clear that sex is puzzle for all organisms humans are presented just as an example.

This is not the first time that Berwick has been caught doctoring a quotation in these pages he also did it in his dismissive review of Richard Dawkins's fine book Climbing Mount Improbable.4 Nor is it the last time. As we shall see, A&B repeatedly fabricate the content of HTMW and misinform their readers about the technical results they marshal to criticize it.

Distortion 1: I uncritically believe that all traits are adaptations.

Let's compare the accusations with what I wrote:

A&B: "In evolutionary terms, every trait we examine is an admixture of physical constraints, natural selection and chance with history. These features are constitutive, not optional. Pinker presents a mutually exclusive conception pitting these factors against each other."

HTMW: "Natural selection should not be pitted against developmental, genetic, or phylogenetic constraints, as if the more important one of them is, the less important the others are" (p. 169). "Organisms can be understood only as interactions among adaptations, by-products of adaptations, and noise" (p. 174).

A&B: "All traits do not lead to a selective advantage, the origin of traits and their maintenance do not demand the same selectionist account. Pinker never embraces this distinction as a vivid part of his adaptationism."

HTMW on all traits leading to selective advantage: "The mind is an adaptation designed by natural selection, but that does not mean that everything we think, feel, and do is biologically adaptive (p. 23)." "The major faculties of the mind . . . show the handiwork of selection. That does not mean that every aspect of the mind is adaptive" (p. 174). "Some readers may be surprised to learn that after seven chapters of reverse-engineering the major parts of the mind, I will conclude by arguing that some of the activities we consider most profound are nonadaptive by-products" (p. 525).

HTMW on origin versus maintenance: "Many organs that we see today have maintained their original function. . . . Others changed their function. . . . [Sometimes] before an organ was selected to assume its current form, it was adapted for something else" (p. 170).

Distortion 2: I am an atomist and a genetic reductionist, reducing every behavior to a simple trait, and then, in a straight line, to a single gene.

This is exactly backwards. Behavior reduced to a single trait?

[The generator of behavior is] the package of information-processing and goal-pursuing mechanisms called the mind.[ . . . ]Any particular deed done today is the effect of dozens of causes. Behavior is the outcome of an internal struggle among many mental modules, and it is played out on the chessboard of opportunities and constraints defined by other people's behavior. (p. 42)

The on-board computers of social organisms, especially of humans, should run sophisticated programs that assess the opportunities and risks at hand and compete or cooperate accordingly. (p. 429)

Behavioral traits reduced to single genes?

The organization of our mental modules comes from our genetic program, but that does not mean that there is a gene for every trait (p. 23).

An emphasis on innate design should not, by the way, be confused with the search for "a gene for" this or that mental organ. [ . . . ]Complex mental organs, like complex physical organs, surely are built by complex genetic recipes, with many genes cooperating in as yet unfathomable ways (pp. 34-35).

Neither adultery nor any other behavior can be in our genes (p. 42).

A&B's specific charges are equally backwards. They charge, "A behavior like 'incest avoidance' gets boiled down to a simple heritable trait." I wrote the opposite: "Incest avoidance showcases the complicated software engineering behind our emotions for other people" (p. 456). A particularly blatant caricature comes in an attempt to educate us on birth order:

Birth order is a clear example of environment, not inheritance, as a prime mover (later-born children could not be expected on average to receive more gullibility genes from their parents). With the wind of adaptationism filling his sails, Pinker must claim that the whole behavioral repertoire response [sic] of first-borns or the inverse behavior of later-borns is all selected for.

Must claim? Here is what I do claim:

[Children] should calculate how to make the best of the hand that nature dealt them and of the dynamics of the poker game they were born into. The historian Frank Sulloway has argued that the elusive nongenetic component of personality is a set of strategies to compete with siblings for parental investment, and that is why children in the same family are so different. Each child develops in a different family ecology and forms a different plan for getting out of childhood alive. (p. 453)

All children being equipped with strategies for competing with siblings is very different from parents' directing gullibility genes to later-borns.

Oddly enough, A&B write as if they themselves believe that genetic atomism and determinism are required by the theory of evolution:

If genes are to serve as accurate bookkeeping chits for maximizing fitness, all the way to love's blushes, then the dotted lines from genes to behaviorsmust run straight and true. Any deviation, any non-determinism or interaction between and among the stepping-stones, and our explanatory hold slips .[emphasis added] (p. 38)

Again, must? It is a commonplace of evolutionary biology that genes are selected because of their probabilistic effects: a gene "for" a trait is a gene that, in comparison with rival alleles, and averaged over environments and over the other genes it appears with, increases the probability that the trait will appear. As A&B acknowledge later in the review, "Biologists have long said a lot about evolution without any detailed knowledge of the steps from genotype to phenotype. All you need are statistical correlations to move from gene to phenotype without a full causal story." Indeed they have. But A&B have just refuted their own claim that "the dotted lines from gene to behavior must run straight and true"!

Distortion 3: HTMW is just story-telling, and presents no basis for evaluating its hypotheses about the biological function of mental faculties.

A&B: Pinker knows that we can always replace one adaptive story by another or add more useful modules as required, but offers us no way to reasonably refrain from doing so.

In fact I "know" no such thing, do "offer us" a way, and do "refrain from doing so" myself.

First, I present standard criteria in biology for evaluating hypotheses about adaptive function. Obviously such criteria must exist. As Ernst Mayr, one of the century's seminal evolutionary biologists, has noted, "The adaptationist question, 'What is the function of a given structure or organ?' has been for centuries the basis of every advance in physiology. If it had not been for the adaptationist program, we probably would still not yet know the functions of thymus, spleen, pituitary, and pineal. Harvey's question 'Why are there valves in the veins?' was a major stepping stone in his discovery of the circulation of blood."

I detail these criteria right from the first chapter (pp. 36-40 also 155-74, 524-26). A good adaptationist explanation specifies a goal relevant to survival or reproduction, the causal structure of the organism's environment, and the engineering designs suited to attain that goal in that environment. It then requires empirical data showing that the trait in question uncannily meets the engineering specs, showing signs of complexity, effectiveness, and specialization in solving the assigned problem, especially in comparison with alternative designs that are biologically possible for that kind of organism.

These criteria are applied throughout. Frequent comparisons with other primates and mammals show that alternatives to what we find in humans, such as lacking language, living in solitude, or mating indiscriminately, are developmentally possible (thus A&B's logical possibility in which genetic variation for such traits is "meager" can be dismissed). And HTMW cites close to 300 empirical reports or literature reviews in evaluating specific hypotheses.

A&B conceal this from the reader. For example, they write, "Are wicked step-parents and parent-child conflict the natural outcome of genetic-payoff investment calculations? HTMW relies on readers' willingness to supply examples from their own lives . . ." No, HTMW relies on cross-cultural and ethnographic surveys, studies of of stepparents in the United States, and national statistics on child abuse and filicide (pp. 433-34).

Here is a particularly unfortunate example:

"There are many stories we tell ourselves. HTMW presents one that allows Pinker to . . . rationalize his particular view of relationships. Read as autobiography this may provide some insight, but as storytelling there are certainly more interesting organizing myths.

A&B do not explain my personal views of relationships, my autobiographical history, or how they claim to know them their innuendo is another attempt to hide the empirical content of HTMW. The "story" in question is the theory of sexual selection and parental investment, and it is a clear counterexample to the claim that adaptationist hypotheses are sterile and unfalsifiable post hoc stories. Darwin first noticed an asymmetry in mating in much of the animal kingdom: males compete, females choose. In the 1960s and 1970s George Williams, John Maynard Smith, and Robert Trivers provided an elegant explanation in terms of parental investment: whichever sex invests more in offspring becomes a limiting resource for the other, and so the less-investing sex competes, the greater-investing sex chooses. In 1979 Donald Symons amassed a vast set of data on human sexuality that supported this theory, refuted competing adaptationist hypotheses, and made many new predictions about promiscuity, jealousy, mate selection, and physical attractiveness. In the 1980s and 1990s these predictions were tested in laboratory experiments, sociological data, in vivo field studies, surveys of the ethnographic literature, and cross-cultural surveys involving tens of thousands of people in thirty-seven cultures-and largely confirmed (pp. 460-93). A&B make no mention of these empirical tests, preferring snide, ad hominemremarks.

Finally, HTMW refutes A&B's cliche that "in adaptationist history/fictions that Pinker fancies there is no end to plausible story telling . . . we can always replace one adaptive story by another or add more modules as required, but [Pinker] offers us no way to reasonably refrain from doing so." A&B do not mention that a major conclusion of HTMW is that many of the most momentous human activities do not meet the criteria for adaptations, including written language, dreams, science, mathematics, music, art, religion, philosophy, and most narrative (pp. 174, 302-06, 340-42, and all of Chapter 8).5

Technical Issues

1. Vision. In the chapter on vision, I suggest that the visual system uses a local, topographic, viewer-centered representation of visible surfaces and their depths, which David Marr called a "2H-D sketch" and which contemporary researchers have modified into what they call a "visual surface representation" (pp. 256-61). A&B claim that "practically no one" believes in it anymore. That is false. Recent overviews by leading perception psychologists such as Berkeley's Stephen Palmer and Harvard's Ken Nakayama rely on a modified 2H-D sketch HTMW's discussion was based on a 1995 synthesis by Nakayama whose title tells a story: "Visual surface representation: A critical link between lower-level and higher-level vision." A&B cite "scientists at MIT" as disproving the visual-surface representation when I showed two of them A&B's claim, the responses were: "To say 'no one believes in it' sounds overly strong. My sense is that surfaces of this sort remain quite popular, both in computer vision and in human vision," and "The claim you cite below is very misleading and exaggerated!"6

2. Behavioral Genetics. I report a near-consensus among human behavioral geneticists: that "much of the variation in personality-about fifty percent-has genetic causes." A&B assert that this naively confuses causation with correlation. I agree that A&B's definition of heritability-"the tendency of offspring to resemble their parents"-does not show that a trait is genetically influenced. But A&B have caricatured an entire field. The claim that variation in adult personality has substantial genetic causes is based on far more than that: that identical twins are more similar than fraternal twins, that biological siblings are more similar than adopted siblings, that these correlations are much the same whether the siblings are reared together or apart, that they remain even after one statistically controls for every possible contaminating factor ever proposed (including shared placentas), that these findings replicate across labs, decades, and countries, and that in several recent studies the variation has been tied to identifiable genes for neurotransmitters or their receptors or transpoters.7

Perhaps fearful that they cannot make the charge stick, A&B turn to other tactics. First, guilt-by-association: they triumphantly refute a silly nine-year-old quotation about IQ from a newspaper reporter. Second, misreporting: they write that I "evidently" rely on Thomas Bouchard's studies of "25" monozygotic twins. In fact Bouchard and his collaborators have studied over 1200 monozygotic twins (112 of them raised apart), and I cite reviews of converging studies of thousands of twins and adoptees conducted by independent teams in half a dozen countries. And finally, the political card: "Pinker's assertion is simply the authority of modern science pressed into the service of speculative fictions-truly biology as ideology." The latter is another trope from Lewontin, summarizing his theory that the current attention to DNA by molecular biologists is an attempt to root social problems in individual failings and thus preserve the status quo.8

3. Population genetics. A&B claim that HTMW falters by not presenting calculations from population genetics, the mathematical modeling of genetic change in evolution:

HTMW contains nothing--literally not one thing- resembling either evolutionary modeling, explicit fitness calculations, or the basics of population or behavioral genetics.

This is false. HTMW was written for a wide audience and does not present equations or proofs, but I explain many results from the explicit modeling of evolution. These include the fundamental theorem of population genetics (p. 163), models of the rate of evolution (pp. 163-64, 205), reconstruction of phylogeny through DNA comparisons (pp. 202-05), Hamilton's models of kin selection (pp. 398-401, 429-30), Trivers's formulation of reciprocal altruism (pp. 402-03, 502-04), Trivers's calculations on parental investment (pp. 440-42, 463-64), models of the genetic costs of incest (pp. 456-58), Grafen's modeling of sexual selection (pp. 500-01), Maynard Smith & Price's game-theoretic analysis of conflict (pp. 494-95), the game-theoretic analyses of bargains, threats, and promises by Schelling, Frank, and Hirshleifer (pp. 409-12), Rogers's calculations of age- and sex-specific discounting of the future (p. 498), Tooby & Cosmides's calculations on coalitional aggression (p. 514), and computer simulations of the evolution of neural networks (pp. 177-79), eyes (p. 164), and strategies of cooperation (pp. 503-04). Data on fitness are cited in discussions of children's age-specific value to parents (p. 452), incest (p. 456), male parental investment (p. 469), dominance (p. 495), and aggression (p. 510), among others. Behavioral genetics is discussed in pages 20-21, 448-449, and elsewhere.

It is true that I do not back up every hypothesis with mathematical models of changing gene frequencies, but then neither do biologists when they study the evolutionary function of the stomach or the eye. Pointedly, neither does Berwick himself in his own foray into evolutionary psychology, in which he confidently states that Chomsky's latest grammatical theory supports a saltationist account of the evolution of language.9

A&B claim to know of specific results from population genetics that undermineHTMW. First they say that population genetics casts strong doubt on whether adaptations can easily evolve: if genes combine in complex ways to determine phenotypes, then the fittest phenotype may not evolve. Of course, fit phenotypesdo evolve. Living things have improbable, intricately engineered parts: eyes and wings and lungs and immune systems and leaves and DNA repair and the other complex machinery that lets plants and animals stay alive. So the fact that some combinations of genes don't evolve into adapted phenotypes simply shows that organisms are not restricted to those combinations of genes over the evolutionary long run. A&B concede the point late in their review:

In recent conversation, James Crow, our foremost population geneticist, has insisted to us that if there were not some trait independence, evolution would grind to a halt, because any change would change all the traits in an organism and so nothing of lasting substance could be built. There is surely something to this.

There surely is, but they did not have to go to "our foremost population geneticist" to learn it. They could have asked me-or any biologist who treats population genetics as a tool to be applied with care and common sense, rather than as a weapon with which to beat opponents.

A&B's specific charge is that I am unaware of a seldom-cited twenty-year-old paper reporting a mathematical model in which "the entire kin-selection- maximize fitness [sic] edifice collapses." Since kin selection-the evolution of traits that benefit relatives because of their shared genes-is a cornerstone of the modern evolutionary analysis of social relations, this would appear to be a major flaw. But A&B's reporting is highly distorted.10

First of all, A&B are incorrect in saying that "Hamilton's original calculation was done only for haploid-diploid organisms where sisters have only half the usual number of chromosomes this makes the bookkeeping easy, allowing us to conflate genes and genotypes." Or as Berwick put it in his other review ofHTMW in the Los Angeles Times, the "cost-benefit arithmetic doesn't balance the genetic books properly, unless you assume people are either ants or bees." In fact Hamilton published his results in two classic papers: the first analyzed the standard diploid case, the second extended it to the haplodiploid social insects.

Second, the paper by Feldman and Cavalli-Sforza does not say that in their interpretation, the entire edifice of kin selection "collapses." It says only that Hamilton's original results are "model-dependent"-a very different claim. Apparently A&B are unaware that in the ensuing two decades many mathematical biologists have developed plausible models in which kin-directed altruism can be proven to evolve, and without the bizarre parameter values that Feldman & Cavalli-Sforza deduced.11

There is a more general error in A&B's referring to population genetics as "the auto mechanics of evolution" and their implication that those who do not use it at every step are courting elementary errors. Models in population genetics are, by necessity, grossly simplified thought experiments about evolution over the short term. A typical model might posit two genes (compared to our 75,000), each coming in two or three alleles (compared to the astronomical numbers possible over the long term), with their interactions fixed (though the interactions themselves may change over the long term), and model the entire environment and entire phenotype as a single coefficient. As in economics, sophisticated mathematical models may be restricted to artificial toy worlds and have little connection to reality. One reformed population geneticist, a collaborator of Feldman's, has recently written:

It seems, thus, that the repeated failure to apply empirically the results of population genetic models to non-trivial evolutionary problems12< HREF="#12"> stems from an attempt to predict one process (say, long-term evolution) on the basis of a model, corresponding to another process (say, short-term evolution). This long-persisting attempt was based on the postulate, tacitly accepted by most in the field (the author of this note included), that at least qualitatively, the behaviour of the long-term process can be fully understood by extrapolation of the analytically well-defined short-term process. We see that this postulate is mathematically wrong.13

No doubt there is much to criticize in HTMW. But bystanders have a right to an accurate characterization of the book's content and of the relevant technical results. Readers of Ahouse and Berwick's review and of similar attacks on evolutionary psychology should ask: if the approach were really that empty, would these critics need to resort to so many misreportings, caricatures, and other questionable tactics?

1 Robert C. Berwick, The Acquisition of Syntactic Knowledge (Cambridge, Mass.: MIT Press, 1985).

2 Robert C. Berwick, "Syntax Facit Saltum: Computation and the Genotype and Phenotype of Language," Journal of Neurolinguistics 10 (1997): 231-49.

3 See HTMW, pp. 44-58 and 165-74, and my letter to the New York Review of Books, 9 October 1997 also Richard Dawkins, The Extended Phenotype (New York:
Oxford University Press, 1982), Daniel Dennett, Darwin's Dangerous Idea (New York: Simon & Schuster, 1995), and Robert Trivers's chapter in Sociobiology and Human Politics, ed. E. White (New York: D. C. Heath, 1981).

4 See Berwick, "Feeling for the Organism," Boston Review 21, no. 6 (December/January 1996), and the responses by Richard Dawkins, Daniel Dennett, and others in the following issue, BR 22, no. 1 (February/March 1997).

5 An irony in A&B's accusation about positing modules is that, Berwick, when arguing in the area of research in which we both work, is quite happy to "add more modules as required" in support of his critique of-surprise!-me: "We can show that the Pinker and Maratsos accounts are misguided. We can do a better job of explaining things via the interaction of two components of language, the morphological and the syntactic. By carving things this way, we shall arrive at an explanation where each "module" is simpler than one in which the explanatory burden is shouldered by the syntactic component alone." Robert Berwick and Amy Weinberg, The Grammatical Basis of Linguistic Performance(Cambridge, Mass.: MIT Press, 1984), p. 214.

6 A&B also misreport the Nature letter by Sinha & Poggio: It was Clinton's facial features, not his "head and shoulders," that were grafted onto Gore Sinha & Poggio's conclusions were about head shape, not "global pose and stance" Sinha & Poggio say nothing about a 2H-D representation, because their demonstration has nothing to do with it.

7 Plomin, R., "Environment and Genes: Determinants of Behavior." American Psychologist 44 (1989): 105-11 "Genes and Behavior: A Special Report," Science 17 (1994): 1685-1739 Special issue of Current Directions in Psychological Science on Behavioral Genetics, (October 1997) Dean Hamer, Living with Our Genes, (New York: Doubleday, 1998). The Devlin et al. Natureletter that A&B refer to is highly tendentious and is answered by Thomas Bouchard in a forthcoming issue.

8 Richard Lewontin, Biology as Ideology: The Doctrine of DNA (New York: Harper Collins, 1991).

9 Berwick, "Syntax facit saltum."

10 A&B unsystematically insert Trivers's very different theory of reciprocal altruism into their critique, but kin selection is the theory relevant to Hamilton and to Feldman & Cavalli-Sforza's principal model and to A&B's mention of step-parents that kicks off their discussion.

11 See, for example, B. Charlesworth and E. L. Charnov, "Kin Selection in Age-Structured Populations," Journal of Theoretical Biology 88 (1981): 103-19 P. D. Taylor and S. A. Frank, "How to Make a Kin Selection Model," Journal of Theoretical Biology 180 (1996): 27-37 P. D. Taylor, "Inclusive Fitness Arguments in Genetic Models of Behaviour," Journal of Mathematical Biology 34 (1996): 654-674 and A. Grafen, "A Geometric View of Relatedness," Oxford Surveys in Evolutionary Biology 2 (1985): 28-89.

12 See for example Richard C. Lewontin, The Genetic Basis of Evolutionary Change (New York: Columbia University Press, 1974).

13 I. Eshel., "On the Changing Concept of Evolutionary Population Stability as a Reflection of a Changing Point of View in the Quantitative Theory of Evolution." Journal of Mathematical Biology 34 (1996): 485-510.

Berwick and Ahouse reply

Steve Pinker suggests that when stripped of vehemence, we all agree that the brain is not just a uniform gray custard, that "processing" lurks behind the mental, and that we share a naturalistic commitment to biology and evolution, including natural selection. But beyond this catechism, we differ more than Pinker's rhetorical narrowing indicates. He suggests that we really disagree only about "which mental processes are functional adaptations." That is a basic point of contention: we think Pinker is much too quick to explain the evolution of mental processes by assigning them adaptive functions. But many of the other difficulties in How the Mind Works (HTMW) follow from his promiscuous adaptationism: reporting premature consensus where there is none uncritically describing behaviors as genetic, implying that they were optimal solutions to an diffusely described past and thus naturalizing them as inevitable underestimating the sheer difficulty of adaptationist arguments identifying "innate mental faculties" with "traits" and embracing a "Darwinian fundamentalism" that effectively equates evolution with natural selection.

Nothing in Pinker's reply urges us to step back from these conclusions. Does Pinker now demonstrate that any behavioral disposition, even "sexual desire" has a coherent evolutionary history or segregates just like the gene for wrinkled peas? No. Does he now establish that only one selective explanation-his- exists for any (hypothetical) trait, paying attention to the space of selective alternatives, physical possibilities, and empirical contingencies? No. Does he probe the selectional regimes 100,000 years ago? The assumption that genetic equilibrium was attained then, the commandeering pace of cultural evolution since? No, none of these things. We did not ask that he turn his book into a technical treatise-we pointed out that he provides none of the technical details.

Panselectionism and Other Distortions

To us Darwinian fundamentalism is a form of irrationalism that, left un-checked, erodes the very theory of evolution it embraces. We are repeatedly told that functionality, design, and adaptedness are explained by only one known physical mechanism: "Natural selection remains the only theory that explains how adaptive complexity, not just any old complexity, can arise" (p. 162) "Because there are no alternatives, we would almost have to accept natural selection as the explanation of life on this planet even if there were no evidence for it" (p. 162) "[natural selection] alone explains what makes life special" (p. 155, our emphasis). But appeal to this doctrinal ("indispensable") principle is coherent only if there is a criterion for functionality, design, and adaptedness that is independent of the actual existence of traits-otherwise, we can "explain" all adaptations, or their negations, equally well. And to this point we have no such independent criterion. HTMW suggests that "reverse engineering" is the independent measure-but we challenged this underpowered metaphor in our review and Pinker did not respond to this either.

Traits and adaptations are simply not obvious. Is it really so clear, as HTMW says, that "sexual desire" is an adaptation to beget children? We already noted in our review that it is controversial why sex originated-increasing genetic variance? parasite defense? mutation elimination?-let alone sexual desire. We simply do not yet know. Why do we see in three dimensions? The answer, to "see in depth," is a near tautology. And it is probably not to recognize objects, contrary to the "consensus" that is presented in HTMW for the now modified "2H-D sketch."1

On the other hand, take HTMW's "obvious" nonadaptation, music. Darwin conjectured music was adaptive, and devoted a whole section to it in The Descent of Man and Selection in Relation to Sex (pp. 496-501). If the answer is not obvious in cases where the traits are relatively easy to measure, then what of our mental capacities and the more plastic, behavioral repertoires associated with them? Allowing that they may well be adaptive, which we do, is not the same as insisting that virtually all are adaptive and that we know why, which is how it is presented in HTMW.

Pinker's response to our description of his position as panselectionist is to complain about distortion and quote a series of one-line caveats that (supposedly) demonstrate that he is aware of our points, anticipated them, and agrees with us. But consider his principal example of distortion. We quoted him as saying: "Why don't women give virgin birth to daughters who are clones of themselves instead of wasting half their pregnancies on sons who lack the machinery to make grandchildren and are nothing but sperm donors?" He responds that, taken in context, the point of the remark was that "sex is puzzle for all organisms humans are presented just as an example." But this was precisely what we rejected: humans are not a good example for understanding the payoff schedule that could select for sexual reproduction, and neither are any other mammals, from aardvark to zebra (nor would any number of many other clades be a good example, e.g., birds). Our point was that structural constraints on evolution might provide the answer to Pinker's question about sexual reproduction. But the notion that there are structural constraints explaining a trait's state is apparently so remote from Pinker's concept of evolution that even when we raise the possibility, he restates (with emphasis) his misunderstanding.

As to the one-line caveats: We noted in our review that Pinker's text includes such hedges and comments, but they are never taken to heart and will not make the rest of the book go away. He really does have to defend his uncorsetted story-telling (the longest chapter, "Family Values," is his exegesis of evolutionary psychology-inspired morals). So we must ask readers to return to our review and see if the examples we challenge are really so unmotivated by the text. Or consider the passages on elephant trunks and evolution (pages 152-53) or cognitive "closure" and the limits of adaptations (pages 558-65). Though we welcome the caution in Pinker's response, our complaint remains: HTMW is a popularization that lacks the spirit of doubt that suffuses scientific practice.

One kind of caveat deserves special mention. This is his widely quoted defense of choosing not to have children thus subverting evolutionary imperative, "and if my genes don't like it they can jump in a lake." In a book whose unrelenting message is that our behaviors were optimized and branded into our genes on the savanna (100,000 to a million years ago) this sudden embrace of genic independence strikes us as merely diversionary, hardly a constitutive part of the view Pinker is defending, according to which "the biggest influence that parents have on their children is at the moment of conception" (p. 449). While conception is a necessary step on the path to a child, Pinker's claim, to be interesting, must be at least partly causal. This is why we find his objection to our accusation of genetic determinism curious. Here again, he drives the narrative forward by making claims that sound strong and controversial only to turn them bland and incontestable through qualification.

One final point on this topic: Pinker resents our implication that his evolutionary stories reflect no more than his biography. Fair enough. We do not know the details of his life or motivations, and were too quick to suggest that his stories only reflect his biography. We offered this hypothesis after having slogged through the book, and found ourselves unconvinced by his insistence that the conclusions are forced on him by data, unmediated by his predilections. Wherever they come from we still insist that there are richer and more interesting organizing myths.

Structure and Function

For those outside evolutionary biology, the stakes in the debate about the extent of functional adaptation may seem esoteric, so some background may be useful. Arguing from function to structure (the explanation for the heart's structure is its function, pumping blood) is an example of teleological thinking. Reasoning from "final causes" is disreputable when discussing systems that lack intention (primarily because it seems to be future states reaching back to form what will be). No one has a problem with saying a pot looks the way it does because the artisan wanted to make it that way, though even here the medium can push against the intentions of the creator. Darwin salvaged this form of thinking in biology through a clever insight, natural selection. After Darwin, function-centric teleological accounts became a kind of shorthand, able to unroll into a more elaborate hypothesis involving variation, heritability, and selection.

A counter-tradition and useful dialectical tension for functionalism is structuralism, an emphasis on conserved structures.2 To return to an earlier example: the explanation of sexual reproduction in humans need not appeal to the function of sexual reproduction in humans-rather, the explanation may well be a matter of constraints and options earlier in this lineage. The selectionist argument that allows us to speak teleologically about a trait requires that the population have variance in the trait. HTMW claims that all of the uniquely human traits emerged through selection on the savanna. But we do not know the variance in human populations with respect to behavioral traits 100,000 years ago, nor do we know the heritability of these dispositions. Unambiguous function assignment is difficult in any case, and from the distance of over a hundred millennia the data you would use to make tentative assignments is quite meager.

Structuralism faces a symmetrical challenge: identifying structures that are "the same" (homologies). These two perspectives provide an essential tension. In the late 1800s functionalism associated itself with evolution and structuralism was associated (by functionalists) with special creation. This has left an unfortunate residue, for as special creation was dismissed structuralism sank with it. Rather than a creative tension between these positions we have the odd situation where some evolutionists assign notions of historical, developmental, or structural constraints to footnotes and caveats. Fortunately, with the string of successes in developmental genetics in the last decade this is changing.3 The distance between a view that prefers a functional-structural dialectic and one that recognizes function as central and anything else as minor epicycles may help explain the rift between Pinker and us.

In part, the differences are matters of temperament: some are content to study the mind and its properties, others insist that we must first agree that the properties are adaptations (even if they can't deliver on the full account that adaptations demand). To us, studying the properties is the interesting thing, not asserting that they are adaptations and then studying them. Moreover, we think that the realization that brains and neurons are homologous structures has been methodologically much more productive (allowing us to do neurobiology by studying mice, crayfish, flies, octopus, and sea hares).

Finally, what are we to make of the suggestion that we (along with Stephen J. Gould, Steven Rose, and Philip Kitcher) are puppets hanging from Dick Lewontin's old and bitter fingers? Well, it is high praise indeed. If we reflect Lewontin's admirable writings, and this puts us on the dais (or in the dungeon) with other compelling writers, we welcome the association. When Steve Pinker asserts that we are all suspicious of attempts to connect psychology with standard evolutionary biology, he is right! We do not believe that nearly enough is known at present about psychology or evolution to link them. But it is possible to link a caricature of each, and this is what we found in HTMW.

Richard Feynman put it well. Our responsibility is to give all the detail to "help others judge the . . . contribution not just the information that leads to judgment in one particular direction or another." The challenge in science writing, indeed science itself, is to tell a crackling good tale abiding the dramatic unities, while respecting scientific ambiguity, details, and unresolved tensions. Pinker's previous book The Language Instinct succeeded here precisely where HTMW fails, and for exactly this reason: the earlier book drew on a forty-year consensus about generative grammar patiently constructed by Noam Chomsky and colleagues, while HTMW forces one prematurely. So when we wrote "You need not believe a word of it" we meant just that: Evolutionary psychology does not force your hand (or your mind) here. You ought to judge for yourself. Pinker could have helped that judgment by presenting alternative explanations antithetical to his views in enough detail. But he didn't.

1 Pinker missed our point about Marr's 2H-D sketch. Of course some people still "believe" in it-though it is important to clear about just what they believe. All we needed to illustrate was a lack of consensus to make our point that his reporting was skewed. According to David Marr's former colleagues, Profs. Shimon Ullman, Tomaso Poggio, Berthold Horn, and Eric Grimson, who, together with Marr, came up with the notion of 'visible surface representation' known as the "2H-D sketch" it was meant to be an integrated representation for surfaces on the way to object recognition. HTMW's picture on page 260, which depicts the 2H-D sketch as a circle and then writes out separate, distinct slots for depth, slant, tilt, color, and surface identification is misleading. Visual surface reconstruction seemingly does matter for certain low-level visual tasks like figure-ground separation, but not, according to Ullman, Poggio, and Heinrich Buetlehoff, for object recognition, not even for 3-D. Further, there is really no hard biological evidence for such a surface representation-scientists simply have not found neurons that respond to surface orientation regardless of where it's coming from. Evidently, neurons behave differently. Moreover, the 2H-D sketch recorded local surface orientation, not really depth (Berthold Horn at the MIT AI Lab did much of this work at MIT in the late 60s and early 1970s, attracting Marr there in the first place). But our point was not to quibble about these details or history: it was simply to point out that nobody had gotten the integrated representation to work (which of course does not bar somebody else from trying to make it work), and that this integrated representation is not a consensus in the field.

2 R. Amundson, "Typology reconsidered: Two Doctrines on the History of Evolutionary Biology" Biology & Philosophy 13, no. 2 (1998):153-77

3 J. Gerhart and M. Kirschner, Cells, Embryos, and Evolution. (Cambridge: Blackwell Science, 1997) Brian Hall, Evolutionary Developmental Biology(New York: Chapman & Hall, 2nd edition, 1998).

Help fund the next generation of Black journalists, editors, and publishers.

Boston Review&rsquos Black Voices in the Public Sphere Fellowship is designed to address the profound lack of diversity in the media by providing aspiring Black media professionals with training, mentorship, networking opportunities, and career development workshops. The program is being funded with the generous support of Derek Schrier, chair of Boston Review&rsquos board of advisors, the Ford Foundation, and the Carnegie Corporation of New York, but we still have $50,000 left to raise to fully fund the fellowship for the next two years. To help reach that goal, if you make a tax-deductible donation to our fellowship fund through August 31 it will be matched 1:1, up to $25,000&mdashso please act now to double your impact. To learn more about the program and our 2021-2022 fellows, click here.


Introduction

Imprinted genes showing parent-of-origin based patterns of expression were first identified in maize [1] and have since been identified in a variety of flowering plants [2–6]. In plants, imprinted expression is primarily observed in the endosperm, which is a nutritive tissue of the seed that is formed when the diploid central cell is fertilized by one of the two sperm cells delivered by the pollen tube. The central cell is epigenetically distinct from most vegetative cells in the plant due to DNA demethylation targeted primarily to Transposable Elements (TEs) [7–9]. This demethylation acts as a primary imprint that distinguishes the female and the male alleles in the endosperm. Maternal and paternal alleles are further distinguished through differential accumulation of histone modifications such as H3K27me3 [10,11] which often marks the maternal allele of paternally expressed genes (PEGs) while maternally expressed genes (MEGs) often show differences in DNA methylation alone [12].

Imprinting has been studied at the genomic level in many plant species [2–6]. While some important genes have conserved imprinting in many species [13], several studies have observed variable imprinting for other genes, with inconsistent imprinting within genotypes of a single species or across species [14,15]. However, understanding the rate of turnover and the origins of imprinted expression patterns has been challenging due in part to methodological inconsistencies across studies and the limitations of available SNPs for allele calls. In Arabidopsis, applying consistent methods and cutoffs across studies reduces apparent variability in imprinting calls [16,17], however many genes cannot be assessed due to a lack of informative SNPs. A lack of SNPs can be due to identical sequence or unalignable regions resulting from large structural changes or presence-absence variation (PAV) of whole genes or features. In maize, many genes and TEs exhibit PAV among genotypes [18–20]. This limits the ability to use SNP-based allele-specific expression analyses to study imprinting, especially for transposons and variable genes. In this study, we develop an alternative approach that relies upon comparisons of expression in reciprocal crosses to assess the imprinting of both conserved and variable genes and TEs across maize genotypes with whole genome assemblies, revealing imprinting for many transposable elements and variable genic sequences.


References

Appleford NE, Evans DJ, Lenton JR, Gaskin P, Croker SJ, Devos KM, Phillips AL, Hedden P (2006) Function and transcript analysis of gibberellin-biosynthetic enzymes in wheat. Planta 223:568–582. doi:10.1007/s00425-005-0104-0

Ashikari M, Sakakibara H, Lin S, Yamamoto T, Takashi T, Nishimura A, Angeles ER, Qian Q, Kitano H, Matsuoka M (2005) Cytokinin oxidase regulates rice grain production. Science 309:741–745

Barboza L, Effgen S, Alonso-Blanco C, Kooke R, Keurentjes JJB, Koornnef M, Alcázar R (2013) Arabidopsis semidwarfs evolved from independent mutations in GA20ox1, ortholog to green revolution dwarf alleles in rice and barley. Proc Natl Acad Sci USA 110:15818–15823

Beales J, Turner A, Griffiths S, Snape JW, Laurie DA (2007) A pseudo-response regulator is misexpressed in the photoperiod insensitive Ppd-D1a mutant of wheat (Triticum aestivum L.). Theor Appl Genet 115:721–733. doi:10.1007/s00122-007-0603-4

Bednarek J, Boulaflous A, Girousse C, Ravel C, Tassy C, Barret P, Bouzidi MF, Mouzeyar S (2012) Down-regulation of the TaGW2 gene by RNA interference results in decreased grain size and weight in wheat. J Exp Bot 63:5945–5955. doi:10.1093/jxb/err313

Bennett MD, Smith JB (1976) Nuclear DNA amounts in angiosperms. Philos Trans R Soc Lond B 274:227–274. doi:10.1098/rstb.1976.0044

Bhati KK, Alok A, Kumar A, Kaur J, Tiwari S, Pandey AK (2016) Silencing of ABCC13 transporter in wheat reveals its involvement in grain development, phytic acid accumulation and lateral root formation. J Exp Bot 67:4379–4389. doi:10.1093/jxb/erw224

Boden S, Cavanagh C, Cullis BR, Ramm K, Greenwood J, Finnegan EJ, Trevaskis B, Swain SM (2015) Ppd-1 is a key regulator of inflorescence architecture and paired spikelet development in wheat. Nat Plants 1. doi:10.1038/nplants.2014.1016

Borlaug NE (1983) Contributions of conventional plant breeding to food production. Science 219:689–693

Börner A, Worland AJ, Plaschke J, Schumann E, Law CN (1993) Pleiotropic effects of genes for reduced height (Rht) and day-length insensitivity (Ppd) on yield and its components for wheat grown in middle Europe. Plant Breed 111:204–216

Börner A, Plaschke J, Korzun V, Worland AJ (1996) The relationships between the dwarfing genes of wheat and rye. Euphytica 89:69–75

Börner A, Roder M, Korzun V (1997) Comparative molecular mapping of GA insensitive Rht loci on chromosomes 4B and 4D of common wheat (Triticum aestivum L.). Theor Appl Genet 95:1133–1137

Börner A, Korzun V, Malyshev S, Ivandic V, Graner A (1999) Molecular mapping of two dwarfing genes differing in their GA response on chromosome 2 H of barley. Theor Appl Genet 99:670–675

Bortiri E, Chuck G, Vollbrecht E, Rocheford T, Martienssen R, Hake S (2006) ramosa2 encodes a LATERAL ORGAN BOUNDARY domain protein that determines the fate of stem cells in branch meristems of maize. Plant Cell 18:574–585. doi:10.1105/tpc.105.039032

Brocklehurst PA (1977) Factors controlling grain weight in wheat. Nature 266:348–349

Cao S, Xu Q, Cao Y, Qian K, An K, Zhu Y, Binzeng H, Zhao H, Kuai B (2005) Loss-of-function mutations in DET2 gene lead to an enhanced resistance to oxidative stress in Arabidopsis. Physiol Plant 123:57–66. doi:10.1111/j.1399-3054.2004.00432.x

Chandler PM, Harding CA (2013) ‘Overgrowth’ mutants in barley and wheat: new alleles and phenotypes of the ‘Green Revolution’ DELLA gene. J Exp Bot 64:1603–1613. doi:10.1093/jxb/ert022

Chandler PM, Marion-Poll A, Ellis M, Gubler F (2002) Mutants at the Slender1 locus of barley cv Himalaya. Molecular and physiological characterization. Plant Physiol 129:181–190. doi:10.1104/pp.010917

Chang C, Lu J, Zhang HP, Ma CX, Sun G (2015) Copy number variation of cytokinin oxidase gene Tackx4 associated with grain weight and chlorophyll content of flag leaf in common wheat. PLoS One 10:e0145970. doi:10.1371/journal.pone.0145970

Che R, Tong H, Shi B, Liu Y, Fang S, Liu D, Xiao Y, Hu B, Liu L, Wang H, Zhao M, Chu C (2015) Control of grain size and rice yield by GL2-mediated brassinosteroid responses. Nat Plants 2:15195. doi:10.1038/nplants.2015.195

Cho JI, Lee SK, Ko S, Kim HK, Jun SH, Lee YH, Bhoo SH, Lee KW, An G, Hahn TR, Jeon JS (2005) Molecular cloning and expression analysis of the cell-wall invertase gene family in rice (Oryza sativa L.). Plant Cell Rep 24:225–236. doi:10.1007/s00299-004-0910-z

Chono M, Honda I, Zeniya H, Yoneyama K, Saisho D, Takeda K, Takatsuto S, Hoshino T, Watanabe Y (2003) A semidwarf phenotype of barley uzu results from a nucleotide substitution in the gene encoding a putative brassinosteroid receptor. Plant Physiol 133:1209–1219. doi:10.1104/pp.103.026195

Christiansen MW, Holm PB, Gregersen PL (2011) Characterization of barley (Hordeum vulgare L.) NAC transcription factors suggests conserved functions compared to both monocots and dicots. BMC Res Notes 4:302. doi:10.1186/1756-0500-4-302

Christiansen MW, Matthewman C, Podzimska-Sroka D, O’Shea C, Lindemose S, Mollegaard NE, Holme IB, Hebelstrup K, Skriver K, Gregersen PL (2016) Barley plants over-expressing the NAC transcription factor gene HvNAC005 show stunting and delay in development combined with early senescence. J Exp Bot 67:5259–5273. doi:10.1093/jxb/erw286

Cockram J, Jones H, Leigh FJ, O’Sullivan D, Powell W, Laurie DA, Greenland AJ (2007) Control of flowering time in temperate cereals: genes, domestication, and sustainable productivity. J Exp Bot 58:1231–1244. doi:10.1093/jxb/erm042

Dahleen LS, Vander Wal LJ, Franckowiak JD (2005) Characterization and molecular mapping of genes determining semidwarfism in barley. J Hered 96:654–662. doi:10.1093/jhered/esi125

Distelfeld A, Pearce SP, Avni R, Scherer B, Uauy C, Piston F, Slade A, Zhao R, Dubcovsky J (2012) Divergent functions of orthologous NAC transcription factors in wheat and rice. Plant Mol Biol 78:515–524. doi:10.1007/s11103-012-9881-6

Distelfeld A, Avni R, Fischer AM (2014) Senescence, nutrient remobilization, and yield in wheat and barley. J Exp Bot 65:3783–3798. doi:10.1093/jxb/ert477

Divi UK, Krishna P (2009) Brassinosteroid: a biotechnological target for enhancing crop yield and stress tolerance. New Biotechnol 26:131–136

Dobrovolskaya O, Pont C, Sibout R, Martinek P, Badaeva E, Murat F, Chosson A, Watanabe N, Prat E, Gautier N, Gautier V, Poncet C, Orlov YL, Krasnikov AA, Berges H, Salina E, Laikova L, Salse J (2015) FRIZZY PANICLE drives supernumerary spikelets in bread wheat. Plant Physiol 167:189–199. doi:10.1104/pp.114.250043

Dockter C, Hansson M (2015) Improving barley culm robustness for secured crop yield in a changing climate. J Exp Bot 66:3499–3509. doi:10.1093/jxb/eru521

Dockter C, Gruszka D, Braumann I, Druka A, Druka I, Franckowiak J, Gough SP, Janeczko A, Kurowska M, Lundqvist J, Lundqvist U, Marzec M, Matyszczak I, Muller AH, Oklestkova J, Schulz B, Zakhrabekova S, Hansson M (2014) Induced variations in brassinosteroid genes define barley height and sturdiness, and expand the green revolution genetic toolkit. Plant Physiol 166:1912–1927. doi:10.1104/pp.114.250738

Doležel J, Binarová P, Lucretti S (1989) Analysis of nuclear DNA content in plant cells by flow cytometry. Biol Plant 31:113–120

Druka A, Franckowiak J, Lundqvist U, Bonar N, Alexander J, Houston K, Radovic S, Shahinnia F, Vendramin V, Morgante M, Stein N, Waugh R (2011) Genetic dissection of barley morphology and development. Plant Physiol 155:617–627. doi:10.1104/pp.110.166249

Dubcovsky J, Dvorak J (2007) Genome plasticity a key factor in the success of polyploid wheat under domestication. Science 316:1862–1866

Echeverry-Solarte M, Kumar A, Kianian S, Mantovani EE, Simsek S, Alamri MS, Mergoum M (2014) Genome-wide genetic dissection of supernumerary spikelet and related traits in common wheat (Triticum aestivum L.). Plant. Genome 7:1–16. doi:10.3835/plantgenome2014.03.0013

Endress PK (2010) Disentangling confusions in inflorescence morphology: Patterns and diversity of reproductive shoot ramification in angiosperms. J Syst Evol 48:225–239. doi:10.1111/j.1759-6831.2010.00087.x

Evans LT (1998) Feeding the ten billion: plants and population growth. Cambridge University Press, Cambridge

Fischer RA, Byerlee D, Edmeades GO (2009) Can technology deliver on the yield challenge to 2050? FAO Expert Meeting on How to Feed the World in 2050, pp 24–26. Rome

Flintham JE, Gale MD (1982) The Tom Thumb dwarfing gene, Rht3 in wheat I. Reduced preharvest sprouting damage to bread making quality. Theor Appl Genet 62:121–126

Flintham JE, Borner A, Worland AJ, Gale MD (1997) Optimizing wheat grain yield: effects of Rht (gibberellin-insensitive) dwarfing genes. J Agric Sci 128:11–25

Forster BP, Franckowiak JD, Lundqvist U, Lyon J, Pitkethly I, Thomas WT (2007) The barley phytomer. Ann Bot 100:725–733. doi:10.1093/aob/mcm183

Franckowiak JD, Lundqvist U (2012) Descriptions of barley genetic stocks for 2012. Barley Genet Newsl 42:36–173

Fu D, Uauy C, Blechl A, Dubcovsky J (2007) RNA interference for wheat functional gene analysis. Transgenic Res 16:689–701. doi:10.1007/s11248-007-9150-7

Gale MD, Marshall GA (1973) Insensitivity to gibberellin in dwarf wheats. Ann Bot 37:729–735

Gale MD, Youssefian S (1985) Dwarfing genes in wheat. In: Russell GE (ed) Progress in plant breeding. Butterworths, London, pp 1–35

Gasparis S, Orczyk W, Zalewski W, Nadolska-Orczyk A (2011) The RNA-mediated silencing of one of the Pin genes in allohexaploid wheat simultaneously decreases the expression of the other, and increases grain hardness. J Exp Bot 62:4025–4036. doi:10.1093/jxb/err103

Gasparis S, Orczyk W, Nadolska-Orczyk A (2013) Sina and Sinb genes in triticale do not determine grain hardness contrary to their orthologs Pina and Pinb in wheat. BMC Plant Biol 13:190

Gasperini D, Greenland A, Hedden P, Dreos R, Harwood W, Griffiths S (2012) Genetic and physiologycal analysis of Rht8 in bread weat: an alternative source of semi-dwarfism with a reduced sensitivity to brassinosteroids. J Exp Bot 63:4419–4436

Gill KS, Gill BS, Endo TR, Taylor T (1996a) Identification and high-density mapping of gene-rich regions in chromosome group I of wheat. Genetics 144:1883–1891

Gill KS, Gill BS, Endot TR, Boyko EV (1996b) Identification and high-density mapping of gene-rich regions in chromosome group 5 of wheat. Genetics 143:1001–1012

Hanif M, Gao F, Liu J, Wen W, Zhang Y, Rasheed A, Xia X, He Z, Cao S (2016) TaTGW6-A1, an ortholog of rice TGW6, is associated with grain weight and yield in bread wheat. Mol Breed 36:1

He X, Qu B, Li W, Zhao X, Teng W, Ma W, Ren Y, Li B, Li Z, Tong Y (2015) The nitrate-inducible NAC transcription factor TaNAC2-5A controls nitrate response and increases wheat yield. Plant Physiol 169:1991–2005. doi:10.1104/pp.15.00568

Hedden P (2003) The genes of the green revolution. Trends Genet 19:5–9

Helliwell CA, Chandler PM, Poole A, Dennis ES, Peacock WJ (2001) The CYP88A cytochrome P450, ent-kaurenoic acid oxidase, catalyzes three steps of the gibberellin biosynthesis pathway. Proc Natl Acad Sci USA 98:2065–2070. doi:10.1073/pnas.041588998

Higgins JA, Bailey PC, Laurie DA (2010) Comparative genomics of flowering time pathways using Brachypodium distachyon as a model for the temperate grasses. PLoS One 5:e10065. doi:10.1371/journal.pone.0010065

Hong Y, Chen L, Du LP, Su Z, Wang J, Ye X, Qi L, Zhang Z (2014) Transcript suppression of TaGW2 increased grain width and weight in bread wheat. Funct Integr Genomics 14:341–349. doi:10.1007/s10142-014-0380-5

Hou J, Jiang Q, Hao C, Wang Y, Zhang H, Zhang X (2014) Global selection on sucrose synthase haplotypes during a century of wheat breeding. Plant Physiol 164:1918–1929. doi:10.1104/pp.113.232454

Houimli SIM, Denden M, Elhadj SB (2008) Induction of salt tolerance in pepper (Capsicum annuum) by 24-epibrassinolide. EurAsian J BioSci 2:83–90

Houston K, McKim SM, Comadran J, Bonar N, Druka I, Uzrek N, Cirillo E, Guzy-Wrobelska J, Collins NC, Halpin C, Hansson M, Dockter C, Druka A, Waugh R (2013) Variation in the interaction between alleles of HvAPETALA2 and microRNA172 determines the density of grains on the barley inflorescence. Proc Natl Acad Sci USA 110:16675–16680. doi:10.1073/pnas.1311681110

Hu M-J, Zhang H-P, Cao J-J, Zhu X-F, Wang S-X, Jiang H, Wu ZY, Lu J, Chang C, Sun G-L, Ma C-X (2016) Characterization of an IAA-glucose hydrolase gene TaTGW6 associated with grain weight in common wheat (Triticum aestivum L.). Mol Breed 36:25

Ishimaru K, Hirotsu N, Madoka Y, Murakami N, Hara N, Onodera H, Kashiwagi T, Ujiie K, Shimizu B, Onishi A, Miyagawa H, Katoh E (2013) Loss of function of the IAA-glucose hydrolase gene TGW6 enhances rice grain weight and increases yield. Nat Genet 45:707–711. doi:10.1038/ng.2612

Jameson PE, Song J (2015) Cytokinin: a key driver of seed yield. J Exp Bot 67:593–606. doi:10.1093/jxb/erv461

Jia Q, Zhang J, Westcott S, Zhang XQ, Bellgard M, Lance R, Li C (2009) GA-20 oxidase as a candidate for the semidwarf gene sdw1/denso in barley. Funct Integr Genomics 9:255–262. doi:10.1007/s10142-009-0120-4

Jiang Q, Hou J, Hao C, Wang L, Ge H, Dong Y, Zhang X (2011) The wheat (T. aestivum) sucrose synthase 2 gene (TaSus2) active in endosperm development is associated with yield traits. Funct Integr Genomics 11:49–61. doi:10.1007/s10142-010-0188-x

Jiang F, Guo M, Yang F, Duncan K, Jackson D, Rafalski A, Wang S, Li B (2012) Mutations in an AP2 transcription factor-like gene affect internode length and leaf shape in maize. PLoS One 7:e37040. doi:10.1371/journal.pone.0037040

Jiang Y, Jiang Q, Hao C, Hou J, Wang L, Zhang H, Zhang S, Chen X, Zhang X (2015) A yield-associated gene TaCWI, in wheat: its function, selection and evolution in global breeding revealed by haplotype analysis. Theor Appl Genet 128:131–143. doi:10.1007/s00122-014-2417-5

Jones-Rhoades MW, Bartel DP, Bartel B (2006) MicroRNAs and their regulatory roles in plants. Annu Rev Plant Biol 57:19–53

Jost M, Taketa S, Mascher M, Himmelbach A, Yuo T, Shahinnia F, Rutten T, Druka A, Schmutzer T, Steuernagel B, Beier S, Taudien S, Scholz U, Morgante M, Waugh R, Stein N (2016) A homolog of Blade-On-Petiole 1 and 2 (BOP1/2) controls internode length and homeotic changes of the barley inflorescence. Plant Physiol 171:1113–1127. doi:10.1104/pp.16.00124

Kapazoglou A, Tondelli A, Papaefthimiou D, Ampatzidou H, Francia E, Stanca MA, Bladenopoulos K, Tsaftaris AS (2010) Epigenetic chromatin modifiers in barley: IV. The study of barley polycomb group (PcG) genes during seed development and in response to external ABA. BMC Plant Biol 10:73. doi:10.1186/1471-2229-10-73

Kebrom TH, Chandler PM, Swain SM, King RW, Richards RA, Spielmeyer W (2012) Inhibition of tiller bud outgrowth in the tin mutant of wheat is associated with precocious internode development. Plant Physiol 160:308–318. doi:10.1104/pp.112.197954

Kellogg EA, Camara PE, Rudall PJ, Ladd P, Malcomber ST, Whipple CJ, Doust AN (2013) Early inflorescence development in the grasses (Poaceae). Front Plant Sci 4:250. doi:10.3389/fpls.2013.00250

Kesavan M, Song JT, Seo HS (2013) Seed size: a priority trait in cereal crops. Physiol Plant 147:113–120. doi:10.1111/j.1399-3054.2012.01664.x

Khripach V, Zhabinskii V, de Groot A (2000) Twenty years of brassinosteroids: steroidal plant hormones warrant better crops for the XXI century. Ann Bot 86:441–447

Komatsuda T, Pourkheirandish M, He C, Azhaguvel P, Kanamori H, Perovic D, Stein N, Graner A, Wicker T, Tagiri A, Lundqvist U, Fujimura T, Matsuoka M, Matsumoto T, Yano M (2007) Six-rowed barley originated from a mutation in a homeodomain-leucine zipper I-class homeobox gene. Proc Natl Acad Sci USA 104:1424–1429. doi:10.1073/pnas.0608580104

Koppolu R, Anwar N, Sakuma S, Tagiri A, Lundqvist U, Pourkheirandish M, Rutten T, Seiler C, Himmelbach A, Ariyadasa R, Youssef HM, Stein N, Sreenivasulu N, Komatsuda T, Schnurbusch T (2013) Six-rowed spike4 (Vrs4) controls spikelet determinacy and row-type in barley. Proc Natl Acad Sci USA 110:13198–13203

Lawlor DW, Paul MJ (2014) Source/sink interactions underpin crop yield: the case for trehalose 6-phosphate/SnRK1 in improvement of wheat. Front Plant Sci 5:418. doi:10.3389/fpls.2014.00418

Lawrence RJ, Pikaard CS (2003) Transgene-induced RNA interference: a strategy for overcoming gene redundancy in polyploids to generate loss-of-function mutations. Plant J 36:114–121. doi:10.1046/j.1365-313X.2003.01857

Lawrenson T, Shorinola O, Stacey N, Li C, Ostergaard L, Patron N, Uauy C, Harwood W (2015) Induction of targeted, heritable mutations in barley and Brassica oleracea using RNA-guided Cas9 nuclease. Genome Biol 16:258. doi:10.1186/s13059-015-0826-7

Li XP, Lan SQ, Feng YR (2006) Effects on agronomic chracteristics of Rht-B1b, Rht-D1b, Rht-B1c wheat dwarfing genes and theirs combinations in artificial climate chamber. J Hebei Agric Sci 10:14–18

Li Y, Fan C, Xing Y, Jiang Y, Luo L, Sun L, Shao D, Xu C, Li X, Xiao J, He Y, Zhang Q (2011) Natural variation in GS5 plays an important role in regulating grain size and yield in rice. Nat Genet 43:1266–1269

Li H, Chen G, Yan W (2015) Molecular characterization of barley 3 H semi-dwarf genes. PLoS One 10:e0120558. doi:10.1371/journal.pone.0120558

Lima Neto MC, Lobo AK, Martins MO, Fontenele AV, Silveira JA (2014) Dissipation of excess photosynthetic energy contributes to salinity tolerance: a comparative study of salt-tolerant Ricinus communis and salt-sensitive Jatropha curcas. J Plant Physiol 171:23–30. doi:10.1016/j.jplph.2013.09.002

Liu H, Bayer M, Druka A, Russell JR, Hackett CA, Poland J, Ramsay L, Hedley PE, Waugh R (2014) An evaluation of genotyping by sequencing (GBS) to map the Breviaristatum-e (ari-e) locus in cultivated barley. BMC Genomics 15:104

Lu J, Chang C, Zhang HP, Wang SX, Sun G, Xiao SH, Ma CX (2015) Identification of a Novel allele of TaCKX6a02 associated with grain size, filling rate and weight of common wheat. PLoS One 10:e0144765. doi:10.1371/journal.pone.0144765

Lundqvist U, Lundqvist A (1988) Induced intermedium mutants in barley: origin, morphology and inheritance. Hereditas 108:13–26

Lundqvist U, Franckowiak JD, Konishi T (1997) New and revised descriptions of barley genes. Barley Genet Newsl 26:22–516

Ma L, Li T, Hao C, Wang Y, Chen X, Zhang X (2015a) TaGS5-3A, a grain size gene selected during wheat improvement for larger kernel and yield. Plant Biotechnol J:1–12. doi:10.1111/pbi.12492

Ma M, Wang Q, Li Z, Cheng H, Li Z, Liu X, Song W, Appels R, Zhao H (2015b) Expression of TaCYP78A3, a gene encoding cytochrome P450 CYP78A3 protein in wheat (Triticum aestivum L.), affects seed size. Plant J 83:312–325. doi:10.1111/tpj.12896

Mayer KF, Martis M, Hedley PE, Simkova H, Liu H, Morris JA, Steuernagel B, Taudien S, Roessner S, Gundlach H, Kubalakova M, Suchankova P, Murat F, Felder M, Nussbaumer T, Graner A, Salse J, Endo T, Sakai H, Tanaka T, Itoh T, Sato K, Platzer M, Matsumoto T, Scholz U, Dolezel J, Waugh R, Stein N (2011) Unlocking the barley genome by chromosomal and comparative genomics. Plant Cell 23:1249–1263. doi:10.1105/tpc.110.082537

Millet E, Pinthus MJ (1984) The association between grain volume and grain weight in wheat. J Cereal Sci 2:31–35

Mochida K, Yamazaki Y, Ogihara Y (2003) Discrimination of homoeologous gene expression in hexaploid wheat by SNP analysis of contigs grouped from a large number of expressed sequence tags. Mol Genet Genomics 270:371–377. doi:10.1007/s00438-003-0939-7

Mok DW, Mok MC (2001) Cytokinin metabolism and action. Annu Rev Plant Physiol Plant Mol Biol 52:89–118

Mulki MA, Von Korff M (2016) CONSTANS controls floral repression by up-regulating VERNALIZATION2 (VRN-H2) in barley. Plant Physiol 170:325–337. doi:10.1104/pp.15.01350

Nadolska-Orczyk A, Gasparis S, Orczyk W (2009) The determinants of grain texture in cereals. J Appl Genet 50:185–197

Nallamilli BR, Zhang J, Mujahid H, Malone BM, Bridges SM, Peng Z (2013) Polycomb group gene OsFIE2 regulates rice (Oryza sativa) seed development and grain filling via a mechanism distinct from Arabidopsis. PLoS Genet 9:e1003322. doi:10.1371/journal.pgen.1003322

Nelson DE, Repetti PP, Adams TR, Creelman RA, Wu J, Warner DC, Anstrom DC, Bensen RJ, Castiglioni PP, Donnarummo MG, Hinchey BS, Kumimoto RW, Maszle DR, Canales RD, Krolikowski KA, Dotson SB, Gutterson N, Ratcliffe OJ, Heard JE (2007) Plant nuclear factor Y (NF-Y) B subunits confer drought tolerance and lead to improved corn yields on water-limited acres. Proc Natl Acad Sci USA 104:16450–16455. doi:10.1073/pnas.0707193104

Olsen AN, Ernst HA, Leggio LL, Skriver K (2005) NAC transcription factors: structurally distinct, functionally diverse. Trends Plant Sci 10:79–87. doi:10.1016/j.tplants.2004.12.010

Paul P, Awasthi A, Rai AK, Gupta SK, Prasad R, Sharma TR, Dhaliwal HS (2012) Reduced tillering in Basmati rice T-DNA insertional mutant OsTEF1 associates with differential expression of stress related genes and transcription factors. Funct Integr Genomics 12:291–304. doi:10.1007/s10142-012-0264-5

Pearce S, Saville R, Vaughan SP, Chandler PM, Wilhelm EP, Sparks CA, Al-Kaff N, Korolev A, Boulton MI, Phillips AL, Hedden P, Nicholson P, Thomas SG (2011) Molecular characterization of Rht-1 dwarfing genes in hexaploid wheat. Plant Physiol 157:1820–1831. doi:10.1104/pp.111.183657

Peng J, Richards DE, Hartley NM, Murphy GP, Devos KM, Flintham JE, Beales J, Fish LJ, Worland AJ, Pelica F (1999) ‘Green Revolution’ genes encode mutant gibberellin response modulators. Nature 400:256–261

Pfeifer M, Kugler KG, Sandve SR, Zhan B, Rudi H, Hvidsten TR, International Wheat Genome Sequencing Consortium, Mayer KF, Olsen OA (2014) Genome interplay in the grain transcriptome of hexaploid bread wheat. Science 345:1250091

Poursarebani N, Seidensticker T, Koppolu R, Trautewig C, Gawronski P, Bini F, Govind G, Rutten T, Sakuma S, Tagiri A, Wolde GM, Youssef HM, Battal A, Ciannamea S, Fusca T, Nussbaumer T, Pozzi C, Borner A, Lundqvist U, Komatsuda T, Salvi S, Tuberosa R, Uauy C, Sreenivasulu N, Rossini L, Schnurbusch T (2015) The genetic basis of composite spike form in barley and ‘Miracle-Wheat’. Genetics 201:155–165. doi:10.1534/genetics.115.176628

Qi W, Sun F, Wang Q, Chen M, Huang Y, Feng YQ, Luo X, Yang J (2011) Rice ethylene-response AP2/ERF factor OsEATB restricts internode elongation by down-regulating a gibberellin biosynthetic gene. Plant Physiol 157:216–228. doi:10.1104/pp.111.179945

Qu B, He X, Wang J, Zhao Y, Teng W, Shao A, Zhao X, Ma W, Wang J, Li B, Li Z, Tong Y (2015) A wheat CCAAT box-binding transcription factor increases the grain yield of wheat with less fertilizer input. Plant Physiol 167:411–423. doi:10.1104/pp.114.246959

Ral JP, Bowerman AF, Li Z, Sirault X, Furbank R, Pritchard JR, Bloemsma M, Cavanagh CR, Howitt CA, Morell MK (2012) Down-regulation of glucan, water-dikinase activity in wheat endosperm increases vegetative biomass and yield. Plant Biotechnol J 10:871–882. doi:10.1111/j.1467-7652.2012.00711.x

Ral JP, Bowerman AF, Li Z, Sirault X, Furbank R, Pritchard JR, Bloemsma M, Cavanagh CR, Howitt CA, Morell M (2013) Down-regulation of glucan, water-dikinase activity in wheat endosperm increase vegetative biomass and yield. Plant Biotechnol J 11:390–391

Ramsay L, Comadran J, Druka A, Marshall DF, Thomas WT, Macaulay M, MacKenzie K, Simpson C, Fuller J, Bonar N, Hayes PM, Lundqvist U, Franckowiak JD, Close TJ, Muehlbauer GJ, Waugh R (2011) INTERMEDIUM-C, a modifier of lateral spikelet fertility in barley, is an ortholog of the maize domestication gene TEOSINTE BRANCHED 1. Nat Genet 43:169–172. doi:10.1038/ng.745

Remizowa MV, Rudall PJ, Choob VV, Sokoloff DD (2013) Racemose inflorescences of monocots: structural and morphogenetic interaction at the flower/inflorescence level. Ann Bot 112:1553–1566. doi:10.1093/aob/mcs246

Saalbach I, Mora-Ramírez I, Weichert N, Andersch F, Guild G, Wieser H, Koehler P, Stangoulis J, Kumlehn J, Weschke W, Weber H (2014) Increased grain yield and micronutrient concentration in transgenic winter wheat by ectopic expression of a barley sucrose transporter. J Cereal Sci 60:75–81. doi:10.1016/j.jcs.2014.01.017

Sakakibara H (2006) Cytokinins: activity, biosynthesis, and translocation. Annu Rev Plant Biol 57:431–449

Sakamoto T, Morinaka Y, Ohnishi T, Sunohara H, Fujioka S, Ueguchi-Tanaka M, Mizutani M, Sakata K, Takatsuto S, Yoshida S, Tanaka H, Kitano H, Matsuoka M (2006) Erect leaves caused by brassinosteroid deficiency increase biomass production and grain yield in rice. Nat Biotechnol 24:105–109

Sakuma S, Pourkheirandish M, Matsumoto T, Koba T, Komatsuda T (2010) Duplication of a well-conserved homeodomain-leucine zipper transcription factor gene in barley generates a copy with more specific functions. Funct Integr Genomics 10:123–133. doi:10.1007/s10142-009-0134-y

Saville RJ, Gosman N, Burt CJ, Makepeace J, Steed A, Corbitt M, Chandler E, Brown JK, Boulton MI, Nicholson P (2012) The ‘Green Revolution’ dwarfing genes play a role in disease resistance in Triticum aestivum and Hordeum vulgare. J Exp Bot 63:1271–1283. doi:10.1093/jxb/err350

Schulte D, Close TJ, Graner A, Langridge P, Matsumoto T, Muehlbauer G, Sato K, Schulman AH, Waugh R, Wise RP, Stein N (2009) The international barley sequencing consortium at the threshold of efficient access to the barley genome. Plant Physiol 149:142–147. doi:10.1104/pp.108.128967

Shahbaz M, Ashraf M, Athar H-u-R (2008) Does exogenous application of 24-epibrassinolide ameliorate salt induced growth inhibition in wheat (Triticum aestivum L.)? Plant Growth Regul 55:51–64. doi:10.1007/s10725-008-9262-y

Shahinnia F, Druka A, Franckowiak J, Morgante M, Waugh R, Stein N (2012) High resolution mapping of Dense spike-ar (dsp.ar) to the genetic centromere of barley chromosome 7H. Theor Appl Genet 124:373–384. doi:10.1007/s00122-011-1712-7

Shan Q, Wang Y, Li J, Gao C (2014) Genome editing in rice and wheat using the CRISPR/Cas system. Nat Protoc 9:2395–2410. doi:10.1038/nprot.2014.157

Sharman BC (1994) Branched heads in wheat and wheat hybrids. Nature 153:497–498

Shaw LM, Turner AS, Laurie DA (2012) The impact of photoperiod insensitive Ppd-1a mutations on the photoperiod pathway across the three genomes of hexaploid wheat (Triticum aestivum). Plant J 71:71–84. doi:10.1111/j.1365-313X.2012.04971.x

Shaw LM, Turner AS, Herry L, Griffiths S, Laurie DA (2013) Mutant alleles of Photoperiod-1 in wheat (Triticum aestivum L.) that confer a late flowering phenotype in long days. PLoS One 8:e79459. doi:10.1371/journal.pone.0079459

Simmonds J, Scott P, Brinton J, Mestre TC, Bush M, Del Blanco A, Dubcovsky J, Uauy C (2016) A splice acceptor site mutation in TaGW2-A1 increases thousand grain weight in tetraploid and hexaploid wheat through wider and longer grains. Theor Appl Genet 129:1099–1112. doi:10.1007/s00122-016-2686-2

Simons KJ, Fellers JP, Trick HN, Zhang Z, Tai YS, Gill BS, Faris JD (2006) Molecular characterization of the major wheat domestication gene Q. Genetics 172:547–555. doi:10.1534/genetics.105.044727

Skov Kristensen P, Dockter C, Lundqvist U, Lu Q, Gregersen PL, Thordal-Christensen H, Hansson M, Pillen K (2016) Genetic mapping of the barley lodging resistance locus Erectoides-k. Plant Breed 135:420–428. doi:10.1111/pbr.12377

Song XJ (2007) A QTL for rice grain width and weight encodes a previously unknown RING-type E3 ubiquitin ligase. Nat Genet 39:623–630

Song J, Jiang L, Jameson PE (2012) Co-ordinate regulation of cytokinin gene family members during flag leaf and reproductive development in wheat. BMC Plant Biol 12:78. doi:10.1186/1471-2229-12-78

Sreenivasulu N, Schnurbusch T (2012) A genetic playground for enhancing grain number in cereals. Trends Plant Sci 17:91–101. doi:10.1016/j.tplants.2011.11.003

Srinivasachary, Gosman N, Steed A, Hollins TW, Bayles R, Jennings P, Nicholson P (2009) Semi-dwarfing Rht-B1 and Rht-D1 loci of wheat differ significantly in their influence on resistance to Fusarium head blight. Theor Appl Genet 118:695–702. doi:10.1007/s00122-008-0930-0

Stephenson TJ, McIntyre CL, Collet C, Xue GP (2007) Genome-wide identification and expression analysis of the NF-Y family of transcription factors in Triticum aestivum. Plant Mol Biol 65:77–92. doi:10.1007/s11103-007-9200-9

Stephenson TJ, McIntyre CL, Collet C, Xue GP (2011) TaNF-YB3 is involved in the regulation of photosynthesis genes in Triticum aestivum. Funct Integr Genomics 11:327–340. doi:10.1007/s10142-011-0212-9

Su Z, Hao C, Wang L, Dong Y, Zhang X (2011) Identification and development of a functional marker of TaGW2 associated with grain weight in bread wheat (Triticum aestivum L.). Theor Appl Genet 122:211–223. doi:10.1007/s00122-010-1437-z

Sun TP, Gubler F (2004) Molecular mechanism of gibberellin signaling in plants. Annu Rev Plant Biol 55:197–223

Tester M, Langridge P (2010) Breeding technologies to increase crop production in a changing world. Science 327:818–822. doi:10.1126/science.1183700

The International Wheat Genome Sequencing C (2014) A chromosome-based draft sequence of the hexaploid bread wheat (Triticum aestivum) genome. Science 345:1251788. doi:10.1126/science.1251788

Travella S, Klimm TE, Keller B (2006) RNA interference-based gene silencing as an efficient tool for functional genomics in hexaploid bread wheat. Plant Physiol 142:6–20. doi:10.1104/pp.106.084517

Turner A, Beales J, Faure S, Dunford RP, Laurie DA (2005) The pseudo-response regulator Ppd-H1 provides adaptation to photoperiod in barley. Science 310:1031–1034. doi:10.1126/science.1117619

Uauy C, Brevis JC, Dubcovsky J (2006a) The high grain protein content gene Gpc-B1 accelerates senescence and has pleiotropic effects on protein content in wheat. J Exp Bot 57:2785–2794. doi:10.1093/jxb/erl047

Uauy C, Distelfeld A, Fahima T, Blechl A, Dubcovsky J (2006b) A NAC gene regulating senescence improves grain protein, zinc, and iron content in wheat. Science 314:1298–1301

Valluru R, Reynolds MP, Salse J (2014) Genetic and molecular bases of yield-associated traits: a translational biology approach between rice and wheat. Theor Appl Genet 127:1463–1489. doi:10.1007/s00122-014-2332-9

Vierstra RD (2009) The ubiquitin-26S proteasome system at the nexus of plant biology. Nat Rev Mol Cell Biol 10:385–397

Vu GT, Wicker T, Buchmann JP, Chandler PM, Matsumoto T, Graner A, Stein N (2010a) Fine mapping and syntenic integration of the semi-dwarfing gene sdw3 of barley. Funct Integr Genomics 10:509–521. doi:10.1007/s10142-010-0173-4

Vu GTH, Dear PH, Caligari PDS, Wilkinson MJ (2010b) BAC-HAPPY mapping (BAP mapping): a new and efficient protocol for physical mapping. PLoS One 5:e9089. doi:10.1371/journal.pone.0009089

Wang J, Yang J, Jia Q, Zhu J, Shang Y, Hua W, Zhou M (2014a) A new QTL for plant height in barley (Hordeum vulgare L.) showing no negative effects on grain yield. PLoS One 9:e90144. doi:10.1371/journal.pone.0090144

Wang Y, Cheng X, Shan Q, Zhang Y, Liu J, Gao C, Qiu JL (2014b) Simultaneous editing of three homoeoalleles in hexaploid bread wheat confers heritable resistance to powdery mildew. Nat Biotechnol 32:947–951. doi:10.1038/nbt.2969

Watanabe K, Breier U, Hensel G, Kumlehn J, Schubert I, Reiss B (2015) Stable gene replacement in barley by targeted double-strand break induction. J Exp Bot 67:1433–1445. doi:10.1093/jxb/erv537

Wei X, Xu J, Guo H, Jiang L, Chen S, Yu C, Zhou Z, Hu P, Zhai H, Wan J (2010) DTH8 suppresses flowering in rice, influencing plant height and yield potential simultaneously. Plant Physiol 153:1747–1758. doi:10.1104/pp.110.156943

Weichert N, Saalbach I, Weichert H, Kohl S, Erban A, Kopka J, Hause B, Varshney A, Sreenivasulu N, Strickert M, Kumlehn J, Weschke W, Weber H (2010) Increasing sucrose uptake capacity of wheat grains stimulates storage protein synthesis. Plant Physiol 152:698–710. doi:10.1104/pp.109.150854

Wen W, Deng Q, Jia H, Wei L, Wei J, Wan H, Yang L, Cao W, Ma Z (2013) Sequence variations of the partially dominant DELLA gene Rht-B1c in wheat and their functional impacts. J Exp Bot 64:3299–3312. doi:10.1093/jxb/ert183

Werner T, Kollmer I, Bartrina I, Holst K, Schmulling T (2006) New insights into the biology of cytokinin degradation. Plant Biol 8:371–381. doi:10.1055/s-2006-923928

Wu CY, Trieu A, Radhakrishnan P, Kwok SF, Harris S, Zhang K, Wang J, Wan J, Zhai H, Takatsuto S, Matsumoto S, Fujioka S, Feldmann KA, Pennell RI (2008) Brassinosteroids regulate grain filling in rice. Plant Cell 20:2130–2145. doi:10.1105/tpc.107.055087

Yadav D, Shavrukov Y, Bazanova N, Chirkova L, Borisjuk N, Kovalchuk N, Ismagul A, Parent B, Langridge P, Hrmova M, Lopato S (2015) Constitutive overexpression of the TaNF-YB4 gene in transgenic wheat significantly improves grain yield. J Exp Bot 66:6635–6650. doi:10.1093/jxb/erv370

Youssefian S, Kirby EJM, Gale MD (1992) Pleiotropic effects of the G.A.-insensitive Rht dwarfing genes in wheat. 1. Effects on development of the ear, stem and leaves. Field Crops Res 28:179–190

Zalewski W, Galuszka P, Gasparis S, Orczyk W, Nadolska-Orczyk A (2010) Silencing of the HvCKX1 gene decreases the cytokinin oxidase/dehydrogenase level in barley and leads to higher plant productivity. J Exp Bot 61:1839–1851. doi:10.1093/jxb/erq052

Zalewski W, Orczyk W, Gasparis S, Nadolska-Orczyk A (2012) HvCKX2 gene silencing by biolistic or Agrobacterium-mediated transformation in barley leads to different phenotypes. BMC Plant Biol 12:206

Zalewski W, Gasparis S, Boczkowska M, Rajchel IK, Kala M, Orczyk W, Nadolska-Orczyk A (2014) Expression patterns of HvCKX genes indicate their role in growth and reproductive development of barley. PLoS One 9:e115729. doi:10.1371/journal.pone.0115729

Zhang R, Lu J, Ma G, Wu Y, Fang M, Zhao Y (1995) Effects of Rht3 gene and 4B-chromosome with Rht3 on photosynthetic carbon-assimilation of wheat. Acta Genet Sin 22:264–271

Zhang L, Zhao YL, Gao LF, Zhao GY, Zhou RH, Zhang BS, Jia JZ (2012) TaCKX6-D1, the ortholog of rice OsCKX2, is associated with grain weight in hexaploid wheat. New Phytol 195:574–584. doi:10.1111/j.1469-8137.2012.04194.x

Zhang Y, Liu J, Xia X, He Z (2014) TaGS-D1, an ortholog of rice OsGS3, is associated with grain weight and grain length in common wheat. Mol Breed 34:1097–1107. doi:10.1007/s11032-014-0102-7

Zheng J, Liu H, Wang Y, Wang L, Chang X, Jing R, Hao C, Zhang X (2014) TEF-7A, a transcript elongation factor gene, influences yield-related traits in bread wheat (Triticum aestivum L.). J Exp Bot 65:5351–5365. doi:10.1093/jxb/eru306


Watch the video: Τι είναι γονική αποξένωση; Σύγκρουση αφοσίωσης. (December 2021).