First printed in R&D Systems' 2004 Catalog.
The use of viruses in cancer research has yielded significant benefits. First, they have provided insight into intracellular signaling and cellular growth control. Second, they have now been established as causative agents in select types of cancer.1, 2
View Larger Image
|Figure 1. Phylogenetic tree of the mouse Wnt genes. [Note: figure adapted from Qian, J. et al. (2003) Genomics 81:34.]|
Up to 15% of all known human cancer may be virally-induced.1 Hepatitis B and C viruses induce hepatocellular carcinoma, Epstein-Barr virus causes Burkitt's lymphoma, and human T cell leukemia virus-1 (HTLV-1) promotes the development of adult Tcell lymphoma.1
Breast cancer in women is a severe epidemiological and medical challenge. It is the leading cause of cancer death in women worldwide and likely will develop in upwards of 250,000 women in the US this year.3 Many causes have been proposed, including diet, hormones, heredity, and benign breast disease. Viruses are also a candidate, and one of the most successful breast cancer experimental models utilizes the interaction of mouse mammary tumor virus (MMTV) with mouse mammary epithelium.4 Over 60 years ago MMTV was identified as a "milk factor," and is now known to be an oncornavirus type B (RNA) retrovirus that utilizes an insertional mutagenesis strategy to infect mammalian cells.2, 5, 6 Almost all types of mouse cells can be infected.4, 5 Only mammary epithelium, however, seems to demonstrate predictable malignant transformation.4 Once infected, MMTV creates a DNA copy of itself and inserts this copy into the cell genome.7, 8 The copy (called a provirus) consists of both coding and noncoding long-term repeat (LTR) sequences, with the latter containing strong gene promoter and enhancer motifs.1 Sites of insertion are generally considered random, and are referred to as integration sites, or "ints." MMTV has been identified in genomic material from mouse chromosomes 1, 4, 6, 7, 15, 16 and 18.7, 9 Thus, there appears to be more than one target for the virus. It is now known that certain sites of integration are also sites of existing genes, either inactive or under strong regulatory control. With proviral insertion, these genes (particularly developmental genes and protooncogenes) become dysregulated and aberrantly expressed (likely due to MMTV LTR promoters and enhancers).10, 11, 12 In some cases, these otherwise silent genes have been named as sequentially numbered ints, with the number reflecting the order of discovery. Int-1 and 4 are developmental proteins on mouse chromosomes 15 and 11 respectively;13, 14 int-2 is FGF-3 on chromosome 7;15 int-3 is also known as Notch-4 on mouse chromosome 17;16 int-5 lies on mouse chromosome 9 and represents the aromatase enzyme that converts androgen to estrogen;17 and int-6 is the p48 protein on chromosome 6 that participates in translation initiation.4, 18 In conclusion, there are a variety of genes, at divergent sites, on multiple chromosomes, that can potentially respond to the promoter mechanisms that accompany MMTV provirus insertions.
At issue, of course, is whether aberrantly expressed molecules actually contribute to tumorigenesis in the breast. In reality, it is perhaps likely that multiple "inappropriate" expressions are necessary for oncogenic transformation.10, 11 In this regard, it has been suggested that the mouse MMTV model may be appropriate for human breast cancer as human tumors express either the mouse mammary tumor virus itself, or a remarkably similar ortholog (98% identity).2, 19
The association of MMTV with the Wnt family of proteins began at the level of int-1. Here, the gene activated on mouse chromosome 15 in the area of int-1 turned out to be a developmentally-regulated, potentially-transforming glycoprotein that bears significant amino acid (aa) orthology to a developmental protein of fruit fly termed wingless.12, 20, 21 The wingless (Wg) gene product is a 469 aa, 53 kDa secreted glycoprotein whose absence creates a "wingless" phenotype in fruit fly.21, 22, 23 When the Wg gene product was compared to known aa sequences, the 370 aa mouse int-1 sequence was found to be most orthologous to Wg (54% aa identity), with an absolute conservation of over 20 cysteine residues.12, 21, 22, 23, 24 With the subsequent cloning of other mouse int-1-related genes, it became clear that a family was emerging, and that an int-1-related designation would be insufficient. Thus, an amalgam was created out of the wingless-int-1 relationship (Wnt; pronounced wint) that would serve as the basis for naming future members.13, 25
This review will concentrate on the human and mouse Wnts. To date, there are at least 19 known Wnts in human, spread over 10 chromosomes, with an additional isoform reported for human Wnt-2b/13 and Wnt-16.26, 27 Multiple identifier numbers exist for many single members, likely due to concurrent or partial cloning. Current numbers will be provided for each individual molecule discussed.
In general, all Wnt genes encode proteins that are destined for secretion, have at least one N-linked glycosylation site, range in length from 350-400 aa, demonstrate 22 to 24 conserved cysteine residues, and show anywhere from 20% to 85% aa identity with other intraspecies Wnts.25, 26, 27, 28 Wnts were initially classified into two groups, based on their ability to either signal through beta-Catenin or initiate calcium flux.29 This was expanded to three classes, based on their ability to transform C57MG cells. Class 1 Wnts (Wnt-1, 2, 3, and 3a) were strong transformers, class 2 Wnts (Wnt-6 and 7) were weak transformers, and class 3 Wnts (Wnt-4, 5a, 5b and 7b) were nontransforming.3031 It has now been proposed that this classification scheme is unfounded, as the prior effects of Wnts may be a function of the receptors involved rather than any intrinsic Wnt-specific characteristics.32
Human Wnt-1 is a 41 kDa, 370 aa protein that contains four potential N-linked glycosylation sites plus 23 cysteines.33, 34 When secreted, it exists as a 343 aa glycoprotein with a molecular weight of 42 to 48 kDa. Twelve of the cysteines lie in the last 70 aa of the molecule, and two of these contribute to potential N-linked glycosylation sites. Almost all studies on Wnt-1 have been done in mouse which shows 99% aa identity to human Wnt-1.12, 20, 34, 35, 36 Here, functional studies have shown that the N-terminus contains receptor binding activity,38 while the C-terminus contributes to cell surface association.39 Although glycosylation is not necessary for activity, the cysteines at positions 116 and 342 are required for transformation.40 Wnt-1 is reported to associate with the cell surface after secretion.36, 37 Whether this is due to heparin/GAG binding is unclear.34, 37, 39, 41 Wnt-1 is generally described as a 44 kDa monomer.36 However, there are also numerous reports of dimer formation.34, 37, 39 As noted, human to mouse, there is 99% aa identity in the mature Wnt-1 segment (one difference over 343 aa).33, 35 Human to rat, there is also 99% aa identity in the mature segment with two aa differences.33, 42 Mammalian cells or cell regions known to express Wnt-1 include spermatids,43 neuroepithelium at the midbrain-hindbrain junction,44 the ventral wall of the midbrain and diencephalon,44 and the dorsal neural tube where it acts laterally on developing somites.45
Full length human Wnt-2 is a 360 aa precursor that contains a 25 aa signal sequence and a 335 aa mature segment. There are 24 cysteine residues and two potential N-linked glycosylation sites. The molecule is glycosylated, with 2 kDa of carbohydrate contributing to the 35 kDa secreted form.46, 47 Mature mouse and rat Wnt-2 are each 96% aa identical to human Wnt-2, differing by 12 and 14 (out of 335) aa, respectively.48, 49 Human Wnt-1 is 40% identical at the aa level to Wnt-2 in the mature region. Although mouse and human Wnt-2 are highly related, a potential exists for two isoforms in mouse. One isoform arises from a polymorphism early in the coding region, generating a 35 aa truncated form. The second isoform lacks a definitive signal sequence. Neither isoform is believed to be secreted. However, they are temporally expressed, and may represent a Wnt-2 regulatory mechanism.48 In the embryo, Wnt-2 is found in lung, pericardium and allantois.48, 50 In the adult, Wnt-2 appears in mammary myoepithelial cells,50 ovarian granulosa cells,51 endometrial epithelium,52 breast fibroblasts,53, 54 and fat pad stroma.55
Human Wnt-2 and Wnt-2b are not alternate splice forms of the same gene. The Wnt-2 gene is found on chromosome 7 while the Wnt-2b gene is located on chromosome 1.27 However, the human Wnt-2b gene does show alternate splicing at the N-terminus with the generation of two apparently functional transcripts.56, 57 The gene itself is composed of at least seven coding exons. The two splice forms share the coding sequence for exons 4 through 7, differing in the use of exon 3 for a 391 aa long form (Wnt-2b2), and exons 1 and 2 for a 372 aa short form (Wnt-2b1).56 Exons 4 through 7 generate 332 common residues, so the difference lies in the N-terminal 59 and 42 aa of 2b2 and 2b1, respectively. Each form contains two N-linked glycosylation sites plus 26 cysteines, 25 of which are conserved. To date, no definitive signal sequence has been identified for either form. Thus, overall, there is about 87% aa identity from human 2b2 to 2b1.56, 57 The long form shows 75% aa identity to human Wnt-2. Mouse Wnt-2b has been cloned and found to be the counterpart of the long 2b2 human form. It is 389 aa in length with an undefined signal sequence, 26 cysteines, and two potential glycosylation sites.58 It shows 96% overall aa identity to human Wnt-2b2. A partial rat clone has also been reported, and over the C-terminal 300 aa shows 99% identity to human Wnt-2b.60 Wnt-2b has been found in ovarian cuboidal epithelium,51 primitive streak mesoderm, heart primordia, and the dorsal midline of the di- and mesencephalon.58
Human Wnt-3 is synthesized as a 355 aa precursor that contains a 21 aa signal sequence and a 334 aa mature segment. There are two potential N-linked glycosylation sites and 24 cysteines, twelve of which lie in the last 70 C-terminal aa.60, 61 Mouse Wnt-3 is also 355 aa in length, and shows 99% aa identity to human Wnt-3 (three differences over 334 aa).14 Although only a partial rat sequence has been reported, it appears it will be identical to that of mouse (based on more than 300 aa).62 Cells known to express Wnt-3 are considerable and include premedullary cells of the hair follicle,63 cells of the embryonic oral and internal enamel epithelium,64 Purkinje cells of the cerebellum, neurons of pons and inferior olivary nuclei (where it may play a role in neuron maturation),65 and limb bud ectoderm.66 Notably, an absence of Wnt-3 in mice precludes the formation of any mesoderm.67
Wnt-3a is not an alternate splice form of Wnt-3. As with other paired Wnts (5 and 7 for example), Wnt-3a may have been part of a gene cluster on chromosome 1 that included Wnt-9a/14. This cluster duplicated, forming Wnt-3 and Wnt-9b/14b/15 on chromosome 17.60 Human Wnt-3a is 352 aa in length with a 24 aa signal sequence and a 328 aa mature region.68 It shows 87% identity to human Wnt-3. The mature segment contains 24 cysteines and two N-linked glycosylation sites. Mature mouse Wnt-3a is also 328 aa in length and known to be a secreted glycoprotein of 44 kDa.69 It exhibits 96% aa identity to human Wnt-3a in the mature segment.70 It is possible that Wnt-3a in mouse and human have different expression patterns.68 In mouse, Wnt-3a has been found in primitive streak ectoderm where it is necessary for correct patterning of the vertebrae and somitic cells.71 It is also found at the margin of the developing cerebral cortex, where it likely plays a role in hippocampal formation,72 and in fibroblasts that are surrounded by fibrin degradation products that accompany wound repair.73
Wnt-4 has the distinction of being the first signaling molecule known to influence sex-determination.74 As a mature molecule, human Wnt-4 is a 46 kDa, 329 aa glycoprotein that contains 24 cysteine residues and two possible N-linked glycosylation sites.53, 69, 74 Mouse Wnt-4 is of equal length, and exhibits 99% aa identity in the mature region.35, 74 A possible alternate splice form has been reported that shows deletion of 14 aa at mature positions 83 through 96.35 Relative to human, rat Wnt-4 is also highly conserved, demonstrating 98% aa identity.75 As noted, Wnt-4 plays a role in sex determination. Wnt-4 expression in mesonephric and gonadal mesenchyme promotes the development of the Müllerian duct and blocks the development of Leydig cells.76 In the same region, Wnt-4 expressing mesenchyme induces a mesenchyme-to-epithelium transition that promotes nephron development.77, 78 In the embryonic adrenal cortex mesoderm, Wnt-4 expression contributes to the proper formation of the zona glomerulosa, a region that secretes aldosterone.78 Wnt-4 is also found in virgin mammary epithelium. Progesterone increases Wnt-4 expression, leading to an enhanced side-branching (but not elongation) of mammary ducts. Finally, Wnt-4 is detected in fibroblasts in areas of wound healing where fibrin degradation products are abundant.73
Human Wnt-5a is an approximately 49 kDa, 343 aa secreted glycoprotein that arises from a 365 aa precursor protein.34, 83 It has four potential N-linked glycosylation sites, 24 cysteine residues, and a 10 aa Wnt-1 family (Wnt-1, 2, 5a) signature that consists of a CKCHGVSGSC motif.83 Mature mouse Wnt-5a is 342 aa in length, and differs from the human by two residues, yielding 99% identity. It also differs in the length of the signal sequence, with the human sequence 22 aa in length and the mouse 37 aa in length.35, 83 Notably, the human Wnt-5a gene is on chromosome 3 while the mouse Wnt-5a gene is on chromosome 14 suggesting that some chromosomal rearrangements must have occurred at this locus during evolution.84 As with mouse, mature rat Wnt-5a is also 342 aa in length and 99% identical to human Wnt-5a. It, too, differs by two aa from the human sequence; however, the two differences are distinct from those in mouse.85 Wnt-5a is expressed by endometrial epithelium,52 stomach mesenchyme,86 mammary gland ductal epithelium,84 embryonic fibroblasts.83, 87 It is also found in the AER (apical ectodermal ridge) of the limb buds, where it is secreted and maintains underlying progress zone mesenchyme (embryonic connective tissue) in an undifferentiated state, providing sufficient cell numbers for expansion and differentiation.88, 89 Wnt-5a is also expressed in neural crest cells destined to become melanocytes. As melanoma cells express high levels of Wnt-5a, the molecule may promote migration of both embryonic melanocytes and melanoma cells.90 Finally, Wnt-5a is secreted by both bone marrow stroma and CD34+ progenitor cells. This promotes the formation of CFU-Mix, CFU-GM, and BFU-E from multilineage progenitors.91
Human Wnt-5b is the genetic paralog of Wnt-5a.92 In other words, a recent gene duplication event created the paralogous pairs, Wnt-3/3a, 5/5a, and 7/7a. Wnt-2 and 2b, discussed previously, are considered to be more distantly related and not the result of a recent gene duplication.92 The mature human Wnt-5b protein consists of a 49 kDa, 342 aa glycoprotein that contains 24 cysteines and four potential N-linked glycosylation sites.69, 93 The mouse ortholog to human Wnt-5b is synthesized as a 359 aa precursor that contains a 17 aa signal sequence and a 342 aa mature region.35 When compared to human Wnt-5b, the mouse Wnt-5b mature segment shows conservation of all cysteines and N-linked glycosylation sites, and an overall identity of 95%, with most substitutions occurring at the C-terminus. As with Wnt-5a, mouse Wnt-5b may exist as a homodimer.34 The rat sequence has been tentatively reported,94 and allowing for adjustments to computational analysis, it will be found to be 98% aa identical to the mouse sequence. Cells known to express Wnt-5b include mammary duct epithelium,55 and embryonic islet alpha- and beta-cells.95 It also appears in chondrocytes, where it promotes a transition from a resting epiphyseal cartilage chondrocyte to an actively dividing growth plate chondrocyte.96, 97
Human Wnt-6 is a 46 kDa, 341 aa secreted glycoprotein that contains two potential N-linked glycosylation sites and 24 cysteine residues.69, 98 Mouse Wnt-6 is synthesized as a 364 aa precursor that contains a 23 aa signal sequence plus a 341 aa mature segment. It shows 98% aa identity with human Wnt-6 in the mature region.35 As with Wnt-5a and -5b, Wnt-6 is secreted as a dimer.34 Rat Wnt-6 is also 98% aa identical to human in the mature segment.99 It has been suggested that Wnt-6 is part of an ancestral gene cluster that included Wnt-1 and 10. In the lineage that gave rise to vertebrates, this cluster duplicated, creating a Wnt-1/6/10a group and a Wnt-1/6/10b group. Later, Wnt-1 was lost from the cluster on human chromosome 2, leaving Wnt-6 and 10a, while Wnt-6 was lost from the three-gene cluster on chromosome 12, leaving Wnt-1 and 10b.98, 100 Cells known to express Wnt-6 include early embryonic ectoderm, which promotes the development of neural crest cells.101 It is also expressed by mammary gland fibroblasts and ductal epithelium,55 and it is found throughout the ectoderm of the limb bud.66
Human Wnt-7a has been cloned and found to be a 349 aa precursor with a 31 aa signal sequence and a 318 aa mature segment. The secreted mature form is 48 kDa with three N-linked glycosylation sites and 24 cysteines.69, 102, 103 Mature mouse Wnt-7a is also 318 aa and shows 98% aa identity to human Wnt-7a.35, 104 Wnt-7a is produced by Müllerian duct epithelium and acts on adjacent mesenchyme to induce the MIS-receptor. Subsequent MIS secretion by Sertoli cells activates the MIS-receptor, which initiates Müllerian duct regression.103 Wnt-7a is also produced by cerebellar granule cells. This induces Mossy fiber axon and growth cone remodeling at future synaptic sites.105 Wnt-7a is also expressed by limb bud dorsal ectoderm which promotes a dorsal phenotype (ulnar formation in the upper extremity).88, 106 Finally, Wnt-7a is found in uterine epithelium,52, 107 and in spinal ganglion neurons and outer hairs of the Organ of Corti.108
Human Wnt-7b is synthesized as a 349 aa precursor that contains a 31 aa signal sequence and a 318 aa mature segment.53, 109 The mature region is characterized by the presence of three N-linked glycosylation sites and 24 cysteine residues. There is 80% identity within the mature regions of human Wnt-7a and 7b. Mouse Wnt-7b is also synthesized as a 349 aa precursor, and the mature region shows 99% aa identity to human Wnt-7b.35 Wnt-7b is secreted as a 46 kDa glycoprotein, and evidence suggests it may circulate as a dimer.34, 69 Wnt-7b is expressed in mammary ductal epithelium, where it maintains cells in an uncommitted state,55 in the ectoderm of the early limb bud,66, 88 and in the developing brain stem caudal to rhombomere 1, where it may be involved in neuron specification and/or axon pathfinding.110
Wnt-8a in human is synthesized as a 355 aa precursor with a 19 aa signal sequence and 336 aa mature segment. The mature region contains three potential N-linked glycosylation sites and 24 cysteines.111 There is also a Wnt-family signature of CKCHGVSGSC at residues 161 through 170.111 The gene for mouse Wnt-8a codes for a protein of 354 aa that is processed into a 40 kDa, 335 aa secreted glycoprotein.114 Mouse to human, mature Wnt-8a is 81% identical at the aa level, with most substitutions occurring in the C-terminus. Wnt-8a is found in primitive streak and early ectoderm during gastrulation. Like Wnt-3a, it appears to be indispensible for heart formation and may control the formation of the early mesoderm.112, 113
Human Wnt-8b is synthesized as a 351 aa precursor that is processed into a 22 aa signal sequence and a 329 aa mature region.114, 115, 116 The mature region contains two potential N-linked glycosylation sites and 24 cysteines, and shows the characteristic Wnt-1 family signature motif. Two polymorphisms have been reported, with a Gly to Ala substitution at position 230 and an Arg to Lys substitution at position 284. Human Wnt-8a and 8b are 66% aa identical in the mature regions. Mature mouse Wnt-8b is also 329 aa in length, and shows over 97% aa identity to human Wnt-8b.117, 118 Wnt-8b is expressed in the dorsal midline of the prosencephalon that gives rise to the hippocampus and corpus callosum.117
Human Wnt-9a, also known initially as Wnt-14, may be the human ortholog of Wnt-9 originally isolated in fish.28, 92 The fact that there are two genes with a number 9 designation suggests that an ancestral gene locus underwent duplication, creating a Wnt-3:Wnt-15 complex on chromosome 17 and a Wnt-3a:Wnt-14 complex on chromosome 1.27, 92, 119 Structurally, human Wnt-9a is synthesized as a 365 aa precursor that is processed into a 29 aa signal sequence and a 336 aa mature segment.68 There is only one potential N-linked glycosylation site in the mature region and this is accompanied by 24 cysteines. Mature mouse Wnt-9a is also 336 aa in length and shares 98% aa identity with human Wnt-9a.120 Rat Wnt-9a also appears to be 98% aa identical to human Wnt-9a, but has a three aa insertion in the center of the molecule.121 Although not much is known about Wnt-9a/14 function and expression, it seems to play a pivotal role in joint formation. In chicken, Wnt-9a is expressed by a small stripe of cells that appear in early long bone cartilage. This expression leads to the production of Growth Differentiation Factor 5 (GDF-5), which leads to synovial joint development.122
Wnt-9b has been cloned in both human and mouse. In human, it consists of a 357 aa precursor that contains a 22 aa signal sequence and a 335 aa mature segment.123 In the mature region there are 24 cysteines and one potential N-linked glycosylation site. Amino acid identity between human Wnt-9b and b is 64%, with most differences occurring in the C-terminus. In mouse, mature Wnt-9b is 336 aa in length and shows 94% identity to human Wnt-9b, with conservation of all 24 cysteines and the one N-terminal glycosylation site.119, 124 Wnt-9b is found in embryonic muscle, epidermis, and intestinal, enamel, and salivary gland epithelium.119
The gene for Wnt-10a codes for a 417 aa precursor that contains a 35 aa signal sequence and a 382 aa mature segment. There are two potential N-linked glycosylation sites and 24 cysteines.98, 125 Mouse Wnt-10a is also expressed as a 417 aa precursor with a 35 aa signal sequence and 382 aa mature region.126 Mouse to human, there is greater than 96% aa identity within mature Wnt-10a. Rat Wnt-10a is also 96% aa identical to the human Wnt-10a mature region.127
Human Wnt-10b is synthesized as a 389 aa precursor that contains a 28 aa signal sequence and a 361 aa mature region. Within the mature segment, there are two potential N-linked glycosylation sites and 24 cysteines.125, 128, 129 It would appear that at least one allelic variant exists in the human where an Asp is substituted for a Gly at position 32 of the mature polypeptide.125 Human Wnt-10a and 10b are difficult to align. Overall, their sequences show less than 60% aa identity. Mature mouse Wnt-10b is a 45 kDa, 361 aa glycosylated polypeptide that shares over 97% aa identity with human Wnt-10b.126, 130, 131 Of note, mouse Wnt-10b is alternately spliced and generates a short, 265aa mature variant that lacks a 96 aa sequence from position 115 to 210.126 The significance of this is unknown. As with mouse Wnt-10b, rat Wnt-10b also shows two splice forms. The long form is orthologous to human at almost 98% aa identity.132 Wnt-10b is found in the mesial cells of dental epithelial thickenings and outer dental epithelium,64 in adult bone marrow stromal cells,91 and in the AER where it likely promotes limb growth via cell proliferation.133 It also has two negative regulator functions. In the bone marrow, it inhibits progenitor commitment to the erythroid lineage,91 and in fat, it is a potent inhibitor of adipogenesis.134
Human Wnt-11 is synthesized as a 354 aa precursor that contains a 24 aa signal sequence and a 330 aa mature segment. It is somewhat unusual among Wnts in that its mature region contains five potential N-linked glycosylation sites plus 24 cysteine residues.135, 136 There may be polymorphisms within the gene as three aa variants have been reported at positions 121 (Ala for Thr), 156 (Gly for Arg), and 271 (Ser for Trp).136 Mature mouse Wnt-11 is a 54 kDa, 330 aa secreted glycoprotein that shares over 97% identity with human Wnt-11.127, 137 The gene for rat Wnt-11 also codes for a 354 aa precursor that shows 97% aa identity to human Wnt-11.138 Cells known to express Wnt-11 include embryonic prehypertrophic chondrocytes,97 somitic cells at the dermatome/myotome junction (in mice, not human),127, 135, 137 lung mesenchyme,135 embryonic colonic and esophageal epithelium,86 and mesenchyme of the developing limb.88, 137 It is also found in cells at the tips of ureteric buds where it is hypothesized to participate with Glial cell-derived Neurotrophic Factor (GDNF) in a positive feedback loop that generates ureter branching.135, 139 Finally, both Wnt-5a and Wnt-11 promote the development of erythroid and monocyte cells in avian bone marrow cultures.140
In the human, there are two Wnt-16 isoforms that are the products of alternate start sites with unique promoters.141 The 16b form is generated from a 365 aa precursor that has a 29 aa signal sequence and a 336 aa mature segment.142 The mature molecule shows 24 cysteines and four potential N-linked glycosylation sites.142 The translation start site (or exon 1b) for this form contributes the first 31 aa with a well-defined signal sequence cleavage site (between aa 29 and 30).141 The 16a form utilizes an alternate start site that generates an initial, and unique, 28 aa before merging with the 16b form sequence at position 32. These 28 N-terminal aa, however, have no well defined signal sequence cleavage site, and as such, the length of the molecule is undefined.141 Relative to the 16b form, the 16a form shows two potential aa substitutions at precursor positions 78 (Gly to Arg) and 259 (Thr to Ile).141 Notably, the 16a form is only found in embryonic pancreas, a pattern distinct from the 16b form that is localized to spleen, kidney, heart and brain.141 Mouse Wnt-16 has one less aa in the mature segment when compared to human Wnt-16b. Unlike most other Wnt orthologs, the mouse ortholog of human Wnt-16b is only 92% aa identical, with a loss of one potential N-linked glycosylation site.142 Only a partial sequence is reported for rat. In the absence of the N-terminal 95 amino acids, it would appear that rat Wnt-16 closely parallels the mouse sequence in overall aa identity relative to human.143 Little is known about Wnt-16 in terms of function or detailed expression. However, it may be important in B cell development.142
The lack of availability of functional natural or recombinant Wnt proteins has hampered studies involving this family. Recently, however, these proteins have been isolated and are now becoming available.144 R&D Systems' 2004 Catalog now includes biologically active recombinant mouse Wnt-3a protein (Catalog # 1324-WN), see the Recombinant Proteins sections for more details. The scientific benefits of functional Wnt proteins is reflected in a recent report utilizing Wnt3a purified from mouse. These researchers reveal that the native Wnt-3a molecule is palmitoylated on two cysteine residues and that this modification is essential for its activity.144 This clearly suggests a new beginning in the study of Wnt biochemistry and function.