Molecular Mechanisms of Neurodegeneration

Evidence suggesting that Homo neanderthalensis contributed the H2 MAPT haplotype to Homo sapiens

J. Hardy, A. Pittman, A. Myers, K. Gwinn-Hardy, H.C. Fung, R. de Silva, M. Hutton, J. Duckworth


The tau (MAPT) locus exists as two distinct clades, H1 and H2. The H1 clade has a normal linkage disequilibrium structure and is the only haplotype found in all populations except those derived from Caucasians. The H2 haplotype is the minor haplotype in Caucasian populations and is not found in other populations. It shows no recombination over a region of 2 Mb with the more common H1 haplotype. The distribution of the haplotype and analysis of the slippage of dinucleotide repeat markers within the haplotype suggest that it entered Homo sapiens populations between approx. 10000 and 30000 years ago. However, sequence comparison of the H2 haplotype with the H1 haplotype and with the chimp sequence suggests that the common founder of the H1 and H2 haplotypes was far earlier than this. We suggest that the H2 haplotype is derived from Homo neanderthalensis and entered H. sapiens populations during the co-existence of these species in Europe from approx. 45000 to 18000 years ago and that the H2 haplotype has been under selection pressure since that time, possibly because of the role of this H1 haplotype in neurodegenerative disease.

  • genome
  • H2 haplotype
  • Homo sapiens
  • human genetics
  • MAPT gene
  • Neanderthal man

Homo sapiens and Homo neanderthalensis lineages diverged approx. 500000 years ago [1]. The latter were the sole human colonizers of Europe until approx. 45000 years ago when H. sapiens also came to Europe [2]. H. neanderthalensis was still present in Europe approx. 18000 years ago [3]. There has much debate about whether H. sapiens and H. neanderthalensis ever mated or whether the former merely displaced the latter from Europe over this time frame [4]. From a genetic perspective, this debate has been limited because the only DNA evidence directly available has been mitochondrial [5] and, although this clearly indicates no contribution to the H. sapiens mitochondrial genome, one cannot generalize from this limited information to autosomal loci, particularly for genes that may have been subjected to selection pressure [4].

The tau (MAPT) locus is very unusual. Over a region of approx. 1.8 Mb, there are two haplotype clades in European populations, H1 and H2 [6,7]. In other populations, only the H1 occurs and shows a normal pattern of recombination [8,9]. This locus contains several other genes besides MAPT [7] (see Figure 1). The H2 haplotype shows remarkably little genetic variation and differs from the H1 haplotype in both sequence and in terms of the orientation of several elements of the locus (Figure 2). These differences prevent recombination between the heterologous clades [10,11].

Figure 1 The consensus build (Build 34: July 2003) of the tau haplotype
Figure 2 H2 RP11 clones to Human Assembly Build 34 (July 2003) showing possible points of inversion between the H1 clade and the H2 haplotype

H2 RP11 clones in the extended tau locus were selected from GenBank®, by using H2-specific sequence markers such as the 238 bp deletion in Intron 9 (del-In9) of MAPT or SNPs that are completely in LD with del-In9. Only RP11-300H14, RP11-162014, RP769P22, RP11-374N3, RP11-1070B7 and RP11-94M7 clones have the DNA sequences AC138688, AC127032, BX544879, AC048388, AC139677 and AC021584 respectively in GenBank®. Since AC021584 for RP11-94M7 is a low pass sequence, this clone is not used for the mapping of H2 clones into July 2003 Human Assembly which is H1 by sequence alignment via UCSC Blat Server. In this Figure, the gene order and orientation of some of the genes in the locus for H1 are shown at the bottom as FLJ25168, CRHR1, MAPT, BC020847, NSF (exon 1 to exon 13) – incomplete duplication of full-length NSF gene – and NSF full-length. On top of this is the gene order and orientation of the same genes for H2. At first glance, one may think H2 is the inversion of H1 from point ‘a’ to ‘b’. However, mapping of H2 clones to July 2003 Human Assembly reveals that the difference between H1 and H2 assemblies is much more complicated than such a simple inversion. In this part of the Figure, H1 assembly coordinates are shown in mega bases (Mb). For example, RP11-769P22 is mapped to H1 assembly in the region of 44.578–44.447 Mb. Notice that 44.578 Mb is higher than 44.447 Mb (arrow is from right to left) because H2 MAPT is in inverse order of H1 MAPT at 44.447–44.578 Mb. Clone RP11-374N3 is mapped into H1 assembly in 4 discontinuous sections marked as 44.21–44.18, 63.47–63.46, 44.137–44.131 and 44.84–44.76 Mb. Clone RP11-1070B7 has short overlap at 44.76 Mb with RP11-374N3. Likewise, Clone RP11-162014 overlaps with RP300H14 for section 45.0615–45.05, 44.116–44.137, 63.465–63.468 and 44.84–44.81 Mb, as aligned with dotted vertical bars. The left end of RP11-300H14 clone shown by two pink arrows (45.18–45.0619 and 45.0619–45.05 Mb) is mapped to H1 continuously in the region of 45.18-45.05 Mb. Alternatively, this left end can be mapped to H1 44.96–44.84 and 44.106–44.116 Mb, although the Blat score is not as good. The two alternative mappings suggest that in addition to having ‘b’ as the inversion break point near NSF, perhaps, the break point is inside full length NSF at ‘c’. ‘c’ is more likely if inversion/duplication happened before NSF duplication in H1 or there were genomic integration from non-H. sapiens such as H. neanderthalensis. Of course, we assume no genomic arrangements between 44.69 and 44.578 Mb and between 44.447 and 44.21 Mb. We mentally connected the sequence GAP though the dotted blue line. It is worth mentioning that alternative mappings of H2 to H1 are possible in the sections marked within highly homologous low copy repeats or LCRs, namely 44.06–44.17, 44.77–44.88, 45.04–45.09, 45.56–45.60 and 63.40–63.46 Mb. Orange-red circles are dinucleotide microsatellites used for H2 haplotype dating, with m1 downstream of BC020847, m2 in MAPT, m3 between MAPT and CRHR1 and m1 in full length NSF. Blue, purple and pink arrow colours stand for inversions, duplications, inversion/duplications respectively. RP11 clones are in green boxes. Note also that it is not clear, relative to the chimp sequence, whether it is H1 or H2 that is inverted. The current build of the chimp sequence is built onto the H1 consensus sequence but this may not be the correct orientation.

We have been interested in the MAPT locus because it is a susceptibility locus for diseases with tangles, including progressive supranuclear palsy [6] and corticobasal degeneration [12], and possibly also including Parkinson dementia complex of Guam [13], a devastating epidemic tangle disease which, at one time, was the major cause of death in South Guam, but has now virtually disappeared [1416].

Analysis of the sequences on the H1 and H2 backgrounds, and comparison of these sequences with those of the chimp (Pan troglodytes) sequence show that, while both H1 and H2 sequences are more similar to each other than to the chimp sequence, they do not follow a predictable relationship: at some sequences, the chimp sequence is similar to H1 and at others, it is similar to H2 (see Table 1, and also [17,18]). Thus the H1 and H2 sequences do not follow a precursor–product relationship and one cannot be derived directly from the other, rather both must have been derived independently from a more distant precursor. Logically, therefore their relationship must be as illustrated in Figure 3.

View this table:
Table 1 Polymorphisms in the MAPT locus that differentiate H1 clades from H2 and comparison with the chimp assembly

Ins/del is the defining insertion deletion polymorphism in intron 9 of the MAPT gene and sthQ7R is the saitohin gene polymorphism [18]. All positions are given relative to Build 34 (July 2003) of the human genome.

Figure 3 Parsimony tree of relationships between chimp MAPT locus and H1 and H2 haplotypes

Parsimony tree showing relationship between the saitohin (Q7R) and the ins/del polymorphisms in the MAPT locus indicating that the H1 and H2 variants of these are more likely to have derived from a common founder than that either H1 or H2 is the predecessor of the other. H1s indicates sthQ7, H2s indicates sthR7. The same diagrams could be drawn for the other polymorphisms in Table 1 (see also [17,18]).

We have two means to date the origin of the H2 haplotype: first, by its geographical distribution, which is consistent with its intrusion into the H. sapiens population at around the time of the founding of the European population [8], approx. 30000 years ago and, secondly, by assuming that the founding event was a single incident, by the slippage of microsatellite repeat markers [19,20]. The analysis of the slippage of microsatellite markers is given in Table 2 and is consistent with a last common ancestor in the time range of this founding. In addition, extensive sequencing of H2 homozygotes has yielded very few polymorphisms in contrast with the large number of variants found exclusively in the H1 clade [20a].

View this table:
Table 2 The predicted age of MAPT H2 haplotype

The 13 DNA samples used were from the CEPH diversity panel [25], and all of them are H2 homozygous defined by Ins/del polymorphism in intron 9 of MAPT gene. The ethnicity of the H2 samples included Algerian, Bedouin, Israeli, Pakistani, French, Italian and Mexican. The ages are calculated using the method described in [19]. Markers m1, m2, m3 and m4 are dinucleotide repeats. Primer sequences were m1F, GAGGGACTGGTAAAGGATTT; m1R, AGGCCGGTAAGAGATCAG; m2F, TTCTAATTCAATGCCTGTTGT; m2R, GGCACCTCAACATAATAATACC; m3F, AGTCAGGTTCTCTCAGGC; m3R, TGGGTCCTGCGTAGATTG; m4F, GAGGGTGCCTTTTAGCTCAT; m4R, GGCCCTGACTAAATCTCCC. These microsatellites are marked as orange red circles in Figure 2: H2 RP11 clones to Human Assembly Build 34 (July 2003).

H. sapiens and H. neanderthalensis co-existed in Europe from approx. 45000 to 18000 years ago [3]. We suggest that a plausible explanation for the ingress of the H2 haplotype is that an ancestral H. neanderthalensis allele, consistent with the parsimony tree of Figure 3, entered the European H. sapiens genome during the period of co-existence, and has spread through selection pressure to its current allele frequency of approx. 25% in this population [8]. The H2 allele is protective against two rare diseases: progressive supranuclear palsy and corticobasal degeneration [6,12]. However, MAPT may also be a susceptibility locus for Parkinson's dementia complex of Guam [13], a disease that reached epidemic proportions around the end of the Second World War and was a major cause of death on the island at that time [14,15]. Thus, while we consider neurodegenerative tangle diseases as rare curiosities, it remains possible that these diseases, like kuru [21], can reach prevalences at which they exert strong selection pressures. In this regard, it is worth noting that both von Economo disease (the sleepy sickness epidemic of Parkinsonism after the Spanish flu epidemic of 1919) [22], and subacute sclerosing panencephalitis, a currently rare but frequently fatal complication of measles infection [23], are also diseases with tangle pathology.

Clearly, in the absence of MAPT sequence data from H. neanderthalensis, these suggestions are speculative. However, they suggest that the search for other genetic remnants of H. neanderthalensis should concentrate on regions with low-recombination rates in European populations in which the sequence diversity is great, as it is here. Interestingly, a recent assessment of LD (linkage disequilibrium) across the genome in different populations suggested that the MAPT locus was the longest region of LD in Europeans [24].


  • Molecular Mechanisms of Neurodegeneration: Joint Biochemical Society/Neuroscience Ireland Focused Meeting (and Satellite Symposium) held at O'Reilly Hall, University College Dublin, Republic of Ireland, 14–16 March 2005. Organized by D. Walsh (University College Dublin, Republic of Ireland), V. Campbell (Trinity College Dublin, Republic of Ireland), M. Fitzgibbon (Trinity College Dublin, Republic of Ireland), B. Irvine (Queen's University Belfast, Northern Ireland), T. Lynch (University College Dublin, Republic of Ireland), J. Johnston (Queen's University Belfast, Northern Ireland), C. O'Neill (University College Cork, Republic of Ireland) and M. Farrell (Royal College of Surgeons in Ireland, Dublin, Republic of Ireland). Edited by C. O'Neill and B. Irvine.

Abbreviations: LD, linkage disequilibrium


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 20a.
  22. 21.
  23. 22.
  24. 23.
  25. 24.
  26. 25.
View Abstract