Homeobox Genes

Homeobox genes play fundamental roles in development and evolution. They are perhaps the best examples of key regulators of gene transcription that are at the heart of the genetic circuitry, regulating different pathways. Homeobox genes have been highly conserved throughout evolution and are involved in the genetic control of the body plan, the determination of cell fate, and several other basic developmental processes. Common among the hundreds of different homeobox genes known today is a highly homologous, 180-bp structural motif, the homeobox . The homeobox encodes a 60-amino acid residue polypeptide chain, the homeodomain, that represents the DNA-binding domain of the respective proteins.

1. History of Homeobox Genes

 Homeotic mutations lead to segmental transformations, suggesting that they are involved in the genetic control of a body plan. The first homeotic mutant, bithorax, was discovered in 1915 (1). This Drosophila gene was later shown to be part of a gene cluster termed the bithorax complex (BX-C(  (2) . Single mutations in the bithorax complex induced major modifications in the body plan, for example, generating four-winged flies. In 1983, the first homeobox gene locus, Antennapedia, was isolated (3), which is part of the Antennapedia complex (Antp). Comparative sequence analysis of the first three cloned homeobox genes, Antennapedia, fushi tarazu, and Ultrabithorax, established a common DNA motif, coined the homeobox (4, 5). This motif was not confined to Drosophila, but was also found in vertebrates, including mice and humans. In 1984, the first homeobox gene in the mouse was cloned by cross-hybridization using a Drosophila homeobox probe (6.(

2. Structure and Function of the Homeodomain

The highly conserved homeodomain is typically composed of 60 amino acid residues. Figure 1 shows the amino acid consensus sequence based on 346 different homeodomain sequences (7. (Comparison of the X-ray crystallography structures of the homeodomains of engrailed and other homeobox genes of the yeast mating-type proteins revealed a helix-turn-helix binding motif (8),  suggesting that all homeodomains bind to DNA in the same manner. The three alpha-helical regions are composed of residues 10–22, 28–38 and 42–58 of the homeodomain (see Fig. 1). Helix 3, the recognition a-helix, binds to the major groove of DNA, whereas helices 1 and 2 lie close to each other in an antiparallel orientation outside the double helix. Thus, only a small number of residues in helix 3 (in particular, that at position 50) and in the N-terminal arm are responsible for the specificity of contacts with the DNA. Furthermore, in vitro binding studies with protein extracts of the locus Antennapedia showed that the homeodomain binds as a monomer to its DNA-binding site, dissociated only slowly, with an estimated half-life of 90  min (9). Other examples suggest that homeobox proteins may dimerize. Transcriptional regulation of homeodomain proteins in vivo was first demonstrated for bicoid, interacting with its target gene hunchback (10-12).

Figure 1. The primary and secondary structures of the homeodomain. The homeodomain contains three well-defined a-helices and a more flexible fourth helix. The schematic illustration of the structure represents a composite derived from the structures of the Antennapedia, engrailed, and MATa2 homeodomains. The amino acid consensus sequence is based on 346 homeodomain sequences. For each position, the amino acid encountered most frequently is listed at the top, while other amino acids are listed beneath in decreasing order of frequency of occurrence. Amino acids occurring fewer than 5 times (1.5%) are not shown. The symbols on top indicate highly conserved positions:   are the most highly conserved positions, with only one or two amino acids found at that position. ƒ indicates highly conserved positions with three to five different amino acids found at a particular position, and * indicates conserved positions, with not more than nine different amino acids.

3. HOX Gene Clusters

Homeobox genes are clustered in Drosophila and organized in two separate clusters, the bithorax (BC-C, 2) and Antennapedia (Antp) complexes (13, 14), and are referred to as HOM-C complexes . The genes of these two gene clusters have been termed HOX genes. HOX genes play a critical role in pattern formation. The order of HOX genes within each cluster correlates with the temporal order of expression of these genes, which are expressed along the anteroposterior axis of the embryo. Interestingly, homeobox genes with sequence similarity to Drosophila HOX genes are also clustered in vertebrates. Both humans and mice have four HOX gene clusters, which reside on different chromosomes. They are termed Hoxa to Hoxd in mice and HOXA to HOXD in humans (15). Altogether, 38  genes reside in these four gene clusters, and all are orientated in the same 5′–3′ direction of transcription (16, 17). Furthermore, in mammals the order of HOX genes within a cluster dictates the time of expression.

 4. Homeotic Mutations

The term homeotic mutations actually refers to HOX gene mutations within the homeotic clusters. For example, a homeotic mutation in the Antennapedia gene complex transforms an antenna into a leg. A mutation in the bithorax gene results in a fly with four wings. A mutation in human HOXD13 results in synpolydactyly. Some of the mouse HOX mutations have been shown to lead to only relatively mild phenotypes. This may be due to functional compensation or to different parallel regulatory pathways.

 In a broader sense, homeotic mutations may encompass mutations of all homeobox genes. Homeotic mutations may represent the classic examples of loss-of-function mutations or gain-of-function mutations. It is interesting that missense mutations causing loss-of-function mutations are often concentrated in the homeodomain or in other functionally important parts of the protein. In contrast, nonsense mutations leading to gene truncations are distributed more widely across the genes. Also, a 50% reduction in the gene product (haploinsufficiency) or overexpression may lead to different clinical phenotypes, suggesting that correct dosage is crucial at certain times in development. This was shown for the first time for PAX6, which leads to aniridie when underexpressed and to other severe eye abnormalities when overexpressed (18, 19).

 5. Homeodomain Gene Families

 Using degenerate oligonucleotide probes or amplification by PCR, a wide variety of novel homeobox genes that represent non-HOX homeobox genes have been identified. Some of these “isolated” (“orphan”) homeodomain gene families contain additional highly conserved domains. Examples of these conserved motifs are the paired-, POU-, LIM-, engrailed-, and zinc-finger motifs, which specify the individual classes of homeobox genes. Each of the two independent DNA-binding domains contacts a specific short sequence on the major groove of the DNA double helix. Examples of the homeobox genes with a paired domain include several members of the PAX gene family (PAX 2, 3, 6); POU domain genes are Pit1, Oct1, and unc 86; genes with a LIM motif include the nematode lin-11 and putative human oncogene rhombotin1; the engrailed class of homeobox genes includes EH1-5, and a gene with a zinc-finger motif is the human ATBF1 gene.

A homeobox page on the different homeobox genes is maintained by Thomas R. Bürglin and can be accessed via the Internet (site currently unavailable. (

 6. Homeobox Genes in Disease

 In the last few years, gene disruption experiments in transgenic mice, as well as positional cloning in humans, have defined a number of homeobox genes associated with genetic diseases and congenital syndromes. The first human HOX gene mutation, HOXD13, was described in families with synpolydactyly (webbing and duplication of fingers) (20). Here the homeodomain was still intact, and the affected members of these families had an expansion of 7–10 additional alanine residues. RIEG, a bicoid -related homeobox gene, was shown to be involved in Rieger syndrome, a complex human disorder with dental, ocular, and further abnormalities (21). Just recently, the chick and Xenopus homologues of RIEG (Pitx2) were shown to determine the left–right asymmetry of internal organs in vertebrates (22). SHOX, a paired-related gene, was shown to play a major role in human growth (23), affecting the height in Turner and Léri-Weill syndrome patients, Langer dwarfism, and idiopathic short stature patients. The MSX1/MSX2 family of genes has been found to be mutated in some forms of craniosynostosis (24). PAX2 and PAX6 genes play a role in eye development, and PAX3 in the Waardenburg syndrome, the most common form of inherited deafness (25). Two gene families in particular, Emx and Otx, have been reported to play a role in brain development, and EMX2 mutations, for example, have been associated with schizencephaly (26).

7. Evolutionary Conservation

Because of their fundamental importance in embryonic development, homeobox genes have been highly conserved throughout evolution. Homeobox genes have been found in plants, fungi, insects, and vertebrates. Both the secondary structures and tertiary structures of homeodomains of different species have been highly conserved. The two HOM-C clusters in Drosophila, and the four HOX clusters in mice and humans, arose by gene duplication and divergence from a common ancestral cluster (27). Such a single gene cluster was found in Amphioxus, the closest invertebrate relative of the vertebrates.

 Functional homology between Drosophila and mouse HOX genes has been shown by gene knockout and transgenic mice (28, 29). Most striking are the gene transfer experiments carried out between mouse and Drosophila. They demonstrated, for example, that the mouse HOXb-6 gene can induce antennal legs when ectopically expressed in Drosophila (30). The ability to compare and extrapolate among species as diverse as Drosophila, mice, and humans is also highlighted in experiments with the mutants of eyeless (ey) in Drosophila, small eye (Pax6) in mice, and Aniridia (PAX6) in humans, which share a 90% sequence identity in their homeodomains. The murine Pax6 is not only strongly homologous to the Drosophila PAX6, but both can induce ectopic eye structures on the wings, legs, and antennae by targeted expression (31).

References

1. C. Bridges and T. H. Morgan (1923) Carnegie Inst. Wash. Publ. 327, 1–251. 

2. E. B. Lewis (1978) Nature 276, 565–570. 

3. R. L. Garber, A. Kuroiwa, and W. J. Gehring (1983) EMBO J. 2, 2027–2036. 

4. M. P. Scott and A. J. Weiner (1984) Proc. Natl. Acad. Sci. USA 81, 4115–4119. 

5. W. McGinnis, M. S. Levine, E. Hafen, A. Kuroiwa, and W. J. Gehring (1984) Nature 308, 428–433.

6. W. McGinnis, C. P. Hart, W. J. Gehring, and F. H. Ruddle (1984) Cell 38, 675–680. 

7. T. R. Bürglin (site currently unavailable). 

8. A. Laughon and M. P. Scott (1984) Nature 310, 25–31. 

9. M. Affolter, A. Percival-Smith, M. Müller, W. Leupin, and W. J. Gehring (1990) Proc. Natl. Acad. Sci. USA 87, 4093–4097. 

10. W. Driever and C. Nüsslein-Volhard (1989) Nature 337, 138–143. 

11. W. Driever, G. Thoma, and C. Nüsslein-Volhard (1989) Nature 340, 363–367. 

12. G. Struhl, K. Struhl, and P. M. Macdonald (1989) Cell 57, 1259–1273. 

13. T. C. Kaufman, R. Lewis, and B. Wakimoto (1980) Genetics 27, 309–362. 

14. T. C. Kaufman, M. Seeger, and G. Olsen (1990) Adv. Gen. 27, 309–362. 

15. D. Duboule, ed. (1994) Guidebook to the Homeobox Genes, IRL Press, Oxford. 

16. M. Kessel and P. Gruss (1990) Science 249, 374–379. 

17. R. Krumlauf (1992) BioEssays 14, 245–252. 

18. T. Jordan, I. Hanson, D. Zaletayev, S. Hodgson, J. Prosser, A. Seawright, N. Hastie, and V. van Heyningen (1992) Nat. Gen. 1, 328–332. 

19. Schedl, A. Ross, M. Lee, D. Engelkamp, P. Rashbass, V. van Heyningen, and N. D. Hastie (1996) Cell 86, 71–82. 

20. Y. Muragaki, S. Mundlos, J. Upton and B. R. Olsen (1996) Science 272, 548–551. 

21. E. V. Semina, R. Reiter, N. J. Leysens, W. L. M. Alward, K. W. Small, N. A. Datson, J. Siegel-Bartelt, D. Bierke-Nelson, P. Bitoun, B. U. Zabel, J. C. Carey, and J. C. Murray (1996) Nat. Gen. 14, 392–399. 

22. A. K. Ryan, B. Blumberg, C. Rodriguez-Esteban, S. Yonei-Tamura, K. Tamura, T. Tsukui, J. de la Peña, W. Sabbagh, J. Greenwald, S. Choe, D. P. Norris, E. J. Robertson, R. M. Evans, M. G. Rosenfeld, and J. C. I. Belmonte (1998) Nature 394, 545–551. 

23. E. Rao, B. Weiss, M. Fukami, A. Rump, B. Niesler, A. Mertz, K. Muroya, G. Binder, S. Kirsch, M. Winkelmann, G. Nordsiek, U. Heinrich, M. H. Breuning, M. B. Ranke, A. Rosenthal, T. Ogata, and G. A. Rappold (1997) Nat. Gen. 16, 54–63. 

24. D. Davidson (1995) Trends Gen. 11, 405–411. 

25. D. Engelkamp and V. van Heyningen (1996) Curr. Opin. Gen. Dev. 6, 334–342. 

26. E. Boncinelli (1997) Curr. Opin. Gen. Dev. 7, 331–337. 

27. C. Kenyon (1994) Cell 78, 175–180. 

28. M. Kessel and P. Gruss (1991) Cell 67, 89–104. 

29. H. LeMouellic, Y. Lallemand, and P. Brûlet (1992) Cell 69, 251–264. 

30. J. Malicki, K. Schughart, and W. McGinnis (1990) Cell 63, 961–967. 

31. G. Halder, P. Callaerts, and W. J. Gehring (1995) Science 267, 1788–1792.