November 25, 2005

What was really Aaron's lineage? (Cohen modal haplotype)

One of the first spectacular applications of Y-chromosome testing was the discovery that Jewish Cohanim exhibited a particular Y-chromosome haplotype, called the Cohen Modal Haplotype (CMH). Jewish Cohens are believed to be descended from the priests of biblical Israel by patrilineal descent, and hence their Y chromosomes should be descended from that of Aaron, the first priest.

In retrospect, the first letter announcing this discovery supported its case by showing a spectacular difference between Cohanim and non-Cohanim Jews. The Cohanim had trace frequencies of haplogroup E3b, and a particular DYS19 allele at high frequency. This finding did prove different histories for priests and non-priests. However, it did not prove descent from a single individual because the YAP- DYS19B combination is not a monophyletic lineage.

In another study, the authors discovered a more extended version of the same haplotype (DYS19-14, DYS388-16, DYS390-23, DYS391-10, DYS392-11, DYS393-12) at high frequency in the Lemba of Southern Africa, as well as in Jews. This was interpreted as evidence for a Jewish origin of the Lemba. This seems likely, given the anomalous existence of a Middle-Eastern haplotype and the oral history of the Lemba. However, once again, it was not proven that people belonging to the CMH were all descended from the biblical Aaron.

A small note: if two men have the same haplotype, it does not mean that they are descended from the same ancestor. This is due to the fact that microsatellites defining haplotypes mutate quite fast, so two unrelated men may have the same haplotype by chance. In fact, the probability of such an accident increases as the number of microsatellites decreases.

The Cohen modal haplotype belongs to a Y-chromosome haplogroup called J or HG-9. A haplogroup is defined by a unique event polymorphism, and men who belong to the same haplogroup are indeed descended from a single man. But, in the case of J, that single man lived more than 10,000 years ago, long before the time of Aaron. However, J is split into two lineages that are also more than 10,000 years old: J2 (or Eu9) and J1 (or Eu10).

If the people who have the CMH are always in Eu9 or in Eu10, then the CMH really reflects priestly descent. But, if it is found in both, then by definition the CMH does not in itself reflect priestly descent, because the common ancestor of a Eu9-CMH and a Eu10-CMH lived earlier than 10,000 years ago, i.e., much earlier than the putative time of Aaron's priesthood (~3 thousand years).

This brings us to yet another study (pdf) which discovered a high frequency of the CMH in Jews. According to this study:
The most-frequent haplotype in all three Jewish groups (the CMH [haplotype 159 in the Appendix]) segregated on a Eu 10 background, together with the three modal haplotypes in Palestinians and Bedouin (haplotypes 144, 151, and 166).
This would suggest that the CMH belonged to J1. However, the author only identified that most CMHs belonged to J1. In the appendix of the same study, it is shown that haplotype 108 is also the CMH, but belongs to haplogroup J2.

Admittedly, 22 CMH-bearing Jews from the study belong to J1, and only 3 ones belong to J2, but this clearly demonstrates that the CMH is not in itself evidence of priestly descent. In principle, Aaron could have been either J1 or J2.

More importantly, the occurrence of the CMH in Cohanim overestimates their descent from Aaron, because at least some of the CMH bearers belong to J1 and some to J2, and they can't both be descended from Aaron.

At this stage, it would appear that we could conclude that the "real" priestly lineage was J1+CMH, because J1+CMH is more frequent than J2+CMH in Jews. Things are however not that simple.

In yet another study the authors state that:
By typing a limited number of Italian Cohanim (A. N. unpublished obs.) for the STRs used here, we determined that the Cohen Modal Haplotype (`an important component in the sharing of Ashkenazic and Sephardic Israelite Y chromosomes', Thomas et al. 2000) does indeed belong to network 1.2.
Network 1.2 falls under the J2 haplogroup. Hence it appears that Cohen Jews, as opposed to Jews in general may belong mainly to haplogroup J2.

At present, no new published information is available on this matter. Several studies have now proven that the CMH occurs in both J1 and J2 backgrounds and does not represent a single lineage. Moreover, Jews themselves are split between the J1 and J2 varieties.

More importantly, the CMH was first identified because of its high frequency compared to other haplotypes. The strength of this evidence is diminished by the finding that CMH chromosomes belong to two unrelated lineages. Furthermore, the dating of CMH chromosomes to Aaron's time should be reconsidered, and the molecular variation within J1 and J2-background CMH and its neighbors should be considered separately.

In conclusion, the true genetic identity of Aaron remains elusive.

UPDATE:

I just discovered the supplementary materials of this comprehensive article on Ashkenazi Jewish Y-chromosome variation.

According to my count, 28 of the Jewish CMHs belong to J1 (6.3%) and 25 belong to J2 (5.7%). This casts further doubt to the CMH as Aaron's lineage, and even if the CMH was Aaron's lineage, it raises the question as to whether he belonged to J1 or J2.

UPDATE II:

In Anatolia, the CMH occurs in:

2 x J2*, 1 x J2f1, 5 x J2f* (total 8 x J2 or 1.5%)
7 x J1* and 1 x J1c (total 8 x J1 or 1.5%)

It is noteworthy that in region 6 (south), 4/33=12% belong to the J1-CMH.

UPDATE III:

Among Armenians, the CMH occurs at a frequency of 1.9%, ranging up to 4.4% in the West region.

No comments: