Y Chromosome III: Evolutionary Strata!

The primary model for Y-chromosome degeneration is a decrease in X-Y recombination. Because and X and Y chromosomes are not kept the same by swapping DNA segments with each other, but the X can still recombine with itself in females, the Y chromosome is allowed to degenerate. We will discuss how this all works next time as I still have to do more reading about the subject.

However, we can still discuss what I think is one of the most interesting phenomena I have encountered in evolutionary biology: evolutionary strata. Although strata are more of an X thing, it still relates to the Y enough for me to want to tell you about it, especially since it’s almost a new way of seeing time and history!

Figure 1: Map of homologous genes in the nonrecombining regions of the human X and Y with strata indicated on the X. Source: Lahn & Page (1999).

In 1999, after mapping the 19 known X-Y homologs (from the X-degenerate region), Lahn & Page noticed the map orders of the X and Y chromosomes did not match (Figure 1). Normally homologous chromosomes are nearly identical in this respect, but not only did the chromosomes mismatch in size and not recombine, they also didn’t align correctly!

Lahn & Page then calculated the Ks values of the 19 homologs. Ks values measure the “estimated mean number of synonymous substitutions per synonymous site.” Such calculations rest upon neutral theory – the third base of a codon tends to not cause a change in the amino acid the DNA codes for, i.e., a mutation at the site has no phenotypic effect and thus selection presumably plays no role. Assuming neutral mutations are allowed to occur at whatever the mutation rate is, the number of synonymous mutations positively correlates with the amount of time since divergence.

Table 1: The Ks values of the 19 X-Y homologues calculated by Lahn and Page (1999).

Once the Ks values were calculated, Lahn & Page noticed that the values grouped in four clusters (Table 1) and that the Ks value of a gene pair in one cluster was significantly different from a gene pair in another cluster (P < 0.02). Additionally, the clusters fell in order according to age (Ks value) on the X chromosome (Figure 2), but were scrambled on the Y (Figure 1). Because the nearer a cluster was to the PARs the lower the Ks values were, Lahn and Page suggest that the “newer” clusters are the most recent to undergo recombination suppression. They hypothesize that the oldest cluster was the first to stop recombining and that recombination suppression moved down the chromosome in a step-wise fashion through chromosomal inversions. Thus they called these clusters “evolutionary strata.”

Lahn & Page (1999) then make two predictions that could confirm or refute this hypothesis:

1) The younger strata on the X have the most X-Y homologs since there has not been as much time for the genes to diverge or become completely erased due to deletions.

2) The youngest stratum has the least X-inactivation. This is predicted because once a gene is inactivated on the Y, the X upregulates itself so the male still benefits from the expression of that gene. However, expression from the X is too high for a female and so the gene on one of the X chromosomes in a female becomes inactivated (Lahn & Page, 1999).

Indeed, Lahn & Page found there are more X-Y homologs in the younger strata (as indicated in Table 1) and they escape X-inactivation to a greater degree than the older strata. (The second part was news to me as I had thought X-inactivation was chromosome-wide, but this has been confirmed in additional studies.) However, 19 genes is not that great of a sample size.

Figure 2: Plot of Ks values (Table 1) versus X-chromosome map position for 16 X-Y genes. Adapted from Lahn & Page (1999).

When Skaletsky et al. (2003) fully sequenced the Y chromosome, they calculated the Ks values of an additional 12 X-Y homologs, including an ampliconic gene pair. While the Ks values followed the same trend discovered by Lahn and Page, the boundaries between the younger strata were no longer so distinct (Figure 3).

Skaletsky et al. (2003) offer several possibilities for blurred boundaries. One possibility is that recombination suppression may have occurred in more than four steps, i.e., stratum 3 is actually multiple strata, or local gene order could have

Figure 3: Plot of KS values against X-chromosome map order with 15 additional X-Y genes. Adapted from Skaletsky et al. (2003).

changed within strata. There is also the possibility of X-Y gene conversion depressing KS values (gene conversion will be discussed later). Other possibilities involve potential errors: the sequence of the X-linked genes could be incorrect (a full sequence wasn’t published for another two years) or Ks estimates could be wrong (as evidenced by the large error bars in Figure 3). There were just too many possibilities to know why the blurred boundaries existed.

In 2005, Ross et al.

Figure 4: Sequence similarities between the X and Y at the fourth and fifth strata. The colors can be ignored here.

published the results of the first full sequence of the human X chromosome and confirmed the hypothesis proposed by Lahn and Page (1999). Instead of investigating the blurred boundaries further, Ross et al. suggest that the fourth strata, the youngest, should be split into two strata due to a marked change in sequence similarity between the strata’s X-Y homologs. The break is also evidenced by progressive increases in G+C content and Alu sequences from stratum 4 to stratum 5 to PAR1 (G+C content and Alu sequences are positively correlated with the presence of genes which also fits into the strata model) (Figure 4).

Further studies of evolutionary strata have only confirmed the ideas laid out by Lahn & Page. One study, Kelkar et al. (2009) split the third stratum into two sections, but I do not understand their methods (what are Markov segmentations?)

(Other studies have shown that strata on the X chromosome exist in mice, chickens, and even the diecious plant, Saline latifolia, whose sex chromosomes certainly share no evolutionary history with those of mammals or birds. We will discuss these other organisms in later posts, however.)

The reason I love evolutionary strata is they “unintentionally” allow us to see time and history in a different light. It reminds me of how you can see the change of the seasons in the hockey stick graph. Here, by measuring abstract Ks values, we can “see” recombination suppression (time) march down a chromosome with physical effects. This record has been destroyed on the Y, of course, but the Y’s degeneration is unintentionally but faithfully recorded by its partner, the X.

So the take-home message for today is that the X chromosome can be segmented into distinct regions where recombination was arrested at different times (separated by millions of years, mind you). We will discuss next time how recombination suppression occurs. I just found 3 or 4 new papers to read the other day so it could be longer than usual for me to next post. After that we will pick up the ampliconic regions and discuss what’s so special about them! Hope you can bear with me!

______________________________

Kelkar A, Thakur V, Ramaswamy R, & Deobagkar D (2009). Characterisation of inactivation domains and evolutionary strata in human X chromosome through Markov segmentation. PloS one, 4 (11) PMID: 19946363

Lahn BT, & Page DC (1999). Four evolutionary strata on the human X chromosome. Science (New York, N.Y.), 286 (5441), 964-7 PMID: 10542153

Ross, M.T., D.V. Grafham, A.J. Coffey, R.A. Gibbs, S. Beck, J. Rogers, D.R. Bentley, & et al. (2005). The DNA sequence of the human X chromosome Nature, 434, 325-337 : 10.1038/nature03440

Skaletsky H, Kuroda-Kawaguchi T, Repping S, Wilson RK, Rozen S, Page DC, & et al. (2003). The male-specific region of the human Y chromosome is a mosaic of discrete sequence classes. Nature, 423 (6942), 825-37 PMID: 12815422