Hybridization in human evolution: Insights from other organisms

During the late Pleistocene, isolated lineages of hominins exchanged genes thus influencing genomic variation in humans in both the past and present. However, the dynamics of this genetic exchange and associated phenotypic consequences through time remain poorly understood. Gene exchange across divergent lineages can result in myriad outcomes arising from these dynamics and the environmental conditions under which it occurs. Here we draw from our collective research across various organisms, illustrating some of the ways in which gene exchange can structure genomic/phenotypic diversity within/among species. We present a range of examples relevant to questions about the evolution of hominins. These examples are not meant to be exhaustive, but rather illustrative of the diverse evolutionary causes/consequences of hybridization, highlighting potential drivers of human evolution in the context of hybridization including: influences on adaptive evolution, climate change, developmental systems, sex-differences in behavior, Haldane's rule and the large X-effect, and transgressive phenotypic variation.

Craniomandibular form and body size variation of first generation mouse hybrids: A model for hominin hybridization

Hybridization occurs in a number of mammalian lineages, including among primate taxa. Analyses of ancient genomes have shown that hybridization between our lineage and other archaic hominins in Eurasia occurred numerous times in the past. However, we still have limited empirical data on what a hybrid skeleton looks like, or how to spot patterns of hybridization among fossils for which there are no genetic data. Here we use experimental mouse models to supplement previous studies of primates. We characterize size and shape variation in the cranium and mandible of three wild-derived inbred mouse strains and their first generation (F1) hybrids. The three parent taxa in our analysis represent lineages that diverged over approximately the same period as the human/Neanderthal/Denisovan lineages and their hybrids are variably successful in the wild. Comparisons of body size, as quantified by long bone measurements, are also presented to determine whether the identified phenotypic effects of hybridization are localized to the cranium or represent overall body size changes. The results indicate that hybrid cranial and mandibular sizes, as well as limb length, exceed that of the parent taxa in all cases. All three F1 hybrid crosses display similar patterns of size and form variation. These results are generally consistent with earlier studies on primates and other mammals, suggesting that the effects of hybridization may be similar across very different scenarios of hybridization, including different levels of hybrid fitness. This paper serves to supplement previous studies aimed at identifying F1 hybrids in the fossil record and to introduce further research that will explore hybrid morphologies using mice as a proxy for better understanding hybridization in the hominin fossil record.

Manifestations of SIV-induced ocular pathology in macaque monkeys

Simian immunodeficiency virus has been shown to cause acquired immunodeficiency syndrome in macaque monkeys. Data gathered from clinical examination and fundus photography have shown that the lentivirus is capable of the induction of choroidal lesions and retinal hemorrhages in the macaque. These findings demonstrate the potential value of the macaque monkey eye as a model of the retinal pathology routinely seen in human AIDS patients.

Transient visual loss due to severe anemia in a patient with AIDS

We present a case of a patient with AIDS who developed a profound anemia caused by zidovudine, an important antiretroviral drug. In the setting of concurrent cytomegalovirus retinitis, the anemia produced transient visual loss that resolved with transfusion of red blood cells. Withdrawal of zidovudine resulted in a stable hemoglobin. This case describes an unusual manifestation of severe anemia. Anemia itself is a very common complication of treatment with zidovudine, one of the most commonly used agents in the treatment of AIDS. The relationship of profound anemia to transient visual loss and the role played by zidovudine in anemia in AIDS patients are discussed.

Fractionator analysis shows loss of neurons in the lateral geniculate nucleus of macaques infected with neurovirulent simian immunodeficiency virus

Infection of macaques with neurovirulent strains of simian immunodeficiency virus (SIVmac) is an experimental model for the neurological manifestations of AIDS. Loss of neurons has been reported in the cerebral cortex following immunodeficiency viral infection, but thalamic structures which may contribute to electrophysiological changes and neurological deficits have not been examined. In this study, the lateral geniculate nucleus (LGN) of macaques inoculated with macrophage-tropic, neurovirulent virus SIVmac239 (R71 and 17E) was examined for neuron loss using the optical fractionator method. Estimates of the number of neurons in the P layers of the lateral geniculate nucleus of age-matched control macaques ranged from 1.0 to 1.3 x 10(6), while the number of neurons in SIV infected macaques ranged from 0.8 to 1.1 x 10(6), reflecting neuron loss of up to 28%. Neuron loss was not observed in the magnocellular layer. The total number of glia and glial density were unchanged. Loss of neurons in the lateral geniculate nucleus was correlated with the pattern of neuropathological changes. Neuron loss was most severe in animals with encephalitis concentrated in the brain stem and subcortical white matter and was less apparent in animals with diffuse encephalitis. Neuron loss in the lateral geniculate nucleus did not explain changes observed in the visual evoked potential, which was severely affected in two animals which showed a loss of 24 and 26%, while it was normal in a third animal which showed neuron loss of 28%.

Free radical production by Nd:YAG laser photodisruption

BACKGROUND AND OBJECTIVES

Plasma and cavitation bubble formation during optical breakdown in aqueous media may produce hydroxyl (*OH) radicals. The authors' objectives were to detect *OH produced by a neodymium:yttrium-aluminum-garnet (Nd:YAG) laser photodisruptor and to determine *OH concentration in relation to laser energy.

MATERIALS AND METHODS

*OH was assayed by measuring absorbance of triiodide (I3-) in a potassium iodide (KI) solution exposed to optical breakdown by an Nd:YAG laser. The concentration-dependent reduction of radical production in relation to cystamine concentration was evaluated.

RESULTS

I3- concentration increased linearly with total irradiation energy and decreased exponentially with increasing cystamine concentration. *OH concentration was calculated using extinction coefficients of I3- and chemical equations relating I3- formation to *OH.

CONCLUSIONS

The authors calculated that approximately 4 x 10(-12) moles of *OH are produced in a typical posterior capsulotomy of 100 mJ of total energy. This *OH concentration could produce strand breaks in approximately 0.4% of vitreous hyaluronic acid molecules, but is unlikely to produce clinical effects.

Complement binding on serum-sensitive and serum-resistant transformants of Neisseria gonorrhoeae: effect of presensitization with a non-bactericidal monoclonal antibody

The binding of C3 and C9 on serum sensitive (FA635) and serum resistant (FA638) transformants of serum sensitive Neisseria gonorrhoeae strain F62 was examined. Previous studies showed that these transformants have Protein IAs which are minimally different by proteinase K cleavage and primary structural and peptide mapping and bear LPS which vary slightly on SDS-PAGE. Binding of C3 and C9 on FA635 exceeded binding on FA638 in NHS and in adsorbed NHS. Monoclonal antibody 4G5, which binds to PI on FA638 but not FA635, increases C9 binding on FA638 to levels 3-3.5 fold greater than on FA635 but does not result in killing. The majority of additional 125IC9 deposited on FA638 following presensitization with 4G5 is released from the bacterial surface by trypsin. These results extend our earlier results with N. gonorrhoeae by showing that, although PI monoclonals can lead to substantial deposition of non-bactericidal C5b-9, this C5b-9 is not fully inserted into the gonococcal outer membrane.

Mechanism of action of blocking immunoglobulin G for Neisseria gonorrhoeae

Blocking immunoglobulin G (IgG) inhibits complement-mediated killing of serum-resistant Neisseria gonorrhoeae (GC) in immune human serum. We examined the mechanism of action of blocking IgG. Presensitization of GC with increasing concentrations of blocking IgG or F(ab')2 before incubation with bactericidal antibody and absorbed pooled normal human serum increased consumption and deposition of the third component of human complement (C3) and the ninth component of human complement (C9) but inhibited killing in dose-related fashion. We next showed that blocking IgG or F(ab')2 partially inhibited binding of bactericidal IgG to GC. Also, binding of a monoclonal antibody recognizing GC outer membrane protein PIII was almost completely inhibited by blocking F(ab')2, confirming other work (Rice, P. A., M. R. Tam, and M. S. Blake, manuscript submitted for publication) showing that PIII is a target for blocking antibody. Studies of the C3 deposition site showed that one quarter of the C3 deposited on GC in the presence of blocking IgG bound covalently to the antibody molecule. Finally, 125I-GC constituents with covalently bound C3 were affinity purified on Sepharose bearing antibodies to C3 and identified by sodium dodecyl sulfate polyacrylamide gel electrophoresis. C3 deposition on a 40,000-mol wt surface protein was enhanced six- to ninefold by blocking IgG, which indicates that the site of complement deposition was altered by blocking antibody. These studies show that blocking IgG competes with bactericidal antibody for binding to GC, but enhances rather than blocks complement activation, and leads to complement deposition at new sites that do not result in serum killing.

Monoclonal antibodies directed against gonococcal protein I vary in bactericidal activity

Monoclonal antibodies (Mab) with specificity for protein I (PI) from Neisseria gonorrhoeae (GC) were examined for bactericidal activity. Mab 4G5 (gamma 3), ID3 (gamma 2a), and 1G6 (gamma 2a) bound to surface-exposed epitopes on PI of GC strain R11 (IA serotype) as assessed by co-agglutination and 125I protein A uptake. Mab 2H1 (gamma 3) that were directed against IB serotype strains and Mab 2E9 (gamma 2a) were negative in co-agglutination and protein A uptake assays and served as controls for some experiments. Only 4G5 and 1D3 were bactericidal for R11 when presensitized organisms were incubated in 10% absorbed, pooled normal human serum (PNHS) or 10% hypogammaglobulinemic serum (H gamma S) despite binding of nearly equivalent numbers of 4G5, 1D3, and 1G6 to R11 during presensitization, as assessed by 125I-protein A uptake. These Mab activated complement to a similar extent on GC R11, leading to deposition of 56.4 X 10(3), 61.9 X 1093), and 47.1 X 10(3) molecules of C3/organism during incubation in 10% C8-deficient serum. Deposition occurred almost exclusively via the classical complement pathway. Measurement of complement component C9 binding to R11 during incubation in H gamma S showed 35,700 molecules of C9/organism with 4G5, 32,600 C9/organism with 1D3, and surprisingly, 29,600 C9/organism with 1G6. Eight thousand four hundred molecules of C9/organism bound to 2E9-coated organisms, 6000 C9/organism to 2H1-coated bacteria, and 3600 C9/organism to nonpresensitized organisms. The C5b-9 complex deposited by 4G5 had a different sedimentation profile by sucrose density gradient analysis from the C5b-9 complex deposited by 1G6, consistent with a different molecular configuration of the bound complex. Mab 1G6 and 1D3, but not 2E9 or 2H1, were able to compete with 125I-4G5 for binding to GC R11. A Mab (2E6) directed against protein III of GC competed weakly with 125I-4G5 for binding to GC R11. Mab 1G6, but not 1D3, blocked 4G5-dependent killing in a dose-related fashion. Both 4G5 and IG6 reacted weakly with native PI of GC R11 by immunoblotting, but neither Mab recognized the 34,800 m.w. fragment of PI generated by trypsin and chymotrypsin treatment of outer membranes. In contrast, 2E9 reacted strongly by immunoblot with both native and cleaved PI of GC R11, suggesting binding to buried determinants of PI. These experiments show that Mab directed against identical or closely associated, surface-exposed epitopes on gonococcal PI differ markedly in bactericidal activity, despite leading to deposition of nearly equivalent numbers of C3 and C9 molecules per organism.

Monoclonal antibodies directed against gonococcal protein I vary in bactericidal activity

Monoclonal antibodies (Mab) with specificity for protein I (PI) from Neisseria gonorrhoeae (GC) were examined for bactericidal activity. Mab 4G5 (gamma 3), ID3 (gamma 2a), and 1G6 (gamma 2a) bound to surface-exposed epitopes on PI of GC strain R11 (IA serotype) as assessed by co-agglutination and 125I protein A uptake. Mab 2H1 (gamma 3) that were directed against IB serotype strains and Mab 2E9 (gamma 2a) were negative in co-agglutination and protein A uptake assays and served as controls for some experiments. Only 4G5 and 1D3 were bactericidal for R11 when presensitized organisms were incubated in 10% absorbed, pooled normal human serum (PNHS) or 10% hypogammaglobulinemic serum (H gamma S) despite binding of nearly equivalent numbers of 4G5, 1D3, and 1G6 to R11 during presensitization, as assessed by 125I-protein A uptake. These Mab activated complement to a similar extent on GC R11, leading to deposition of 56.4 X 10(3), 61.9 X 1093), and 47.1 X 10(3) molecules of C3/organism during incubation in 10% C8-deficient serum. Deposition occurred almost exclusively via the classical complement pathway. Measurement of complement component C9 binding to R11 during incubation in H gamma S showed 35,700 molecules of C9/organism with 4G5, 32,600 C9/organism with 1D3, and surprisingly, 29,600 C9/organism with 1G6. Eight thousand four hundred molecules of C9/organism bound to 2E9-coated organisms, 6000 C9/organism to 2H1-coated bacteria, and 3600 C9/organism to nonpresensitized organisms. The C5b-9 complex deposited by 4G5 had a different sedimentation profile by sucrose density gradient analysis from the C5b-9 complex deposited by 1G6, consistent with a different molecular configuration of the bound complex. Mab 1G6 and 1D3, but not 2E9 or 2H1, were able to compete with 125I-4G5 for binding to GC R11. A Mab (2E6) directed against protein III of GC competed weakly with 125I-4G5 for binding to GC R11. Mab 1G6, but not 1D3, blocked 4G5-dependent killing in a dose-related fashion. Both 4G5 and IG6 reacted weakly with native PI of GC R11 by immunoblotting, but neither Mab recognized the 34,800 m.w. fragment of PI generated by trypsin and chymotrypsin treatment of outer membranes. In contrast, 2E9 reacted strongly by immunoblot with both native and cleaved PI of GC R11, suggesting binding to buried determinants of PI. These experiments show that Mab directed against identical or closely associated, surface-exposed epitopes on gonococcal PI differ markedly in bactericidal activity, despite leading to deposition of nearly equivalent numbers of C3 and C9 molecules per organism.

Bactericidal but not nonbactericidal C5b-9 is associated with distinctive outer membrane proteins in Neisseria gonorrhoeae

In this study, we examined the bacterial constituents associated with the complement C5b-9 complex in detergent extracts from serum-treated Neisseria gonorrhoeae (GC). 125I surface-labeled GC were incubated in 10% serum, were washed, and were solubilized in the zwitterionic sulfobetaine detergent SB12. Immunoprecipitation of 125I-GC from the extract with anti-C5 Sepharose was followed by 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis autoradiography of immunoprecipitated material. Polyacrylamide gel analysis of surface-labeled 125I-GC showed prominent bands for proteins I and III for both serum-resistant GC strain 6305 and serum-sensitive GC strain 7189. These same bands were visible with similar intensity in the SB12 extracts from presensitized and non-presensitized 6305 and 7189 after serum incubation. For those organisms bearing bactericidal C5b-9 (6305 + IgG and 7189 +/- IgG), additional distinctive bands immunoprecipitated with antibody to C5 Sepharose. These components were of 93,000, 44,000 40,000, and 15,000 daltons for 6305 + IgG, and were of 90,000, 50,000, 44,000, and 19,000 daltons for 7189 +/- IgG. Nonbactericidal C5b-9 extracted from the surface of 6305 incubated in serum, but not sensitized with antibody, was not associated with these distinctive proteins. However, this nonbactericidal C5b-9 did have a different pattern of associated bacterial surface constituents from that observed in control samples incubated with antibody to human serum albumin, which were similar to those with nonserum-incubated organisms. These studies support our earlier experiments which demonstrated that C5b-9 is in a different molecular configuration on the surface of serum-resistant GC from that on the surface of serum-sensitive GC or resistant GC rendered sensitive with bactericidal antibody.

Bactericidal but not nonbactericidal C5b-9 is associated with distinctive outer membrane proteins in Neisseria gonorrhoeae

In this study, we examined the bacterial constituents associated with the complement C5b-9 complex in detergent extracts from serum-treated Neisseria gonorrhoeae (GC). 125I surface-labeled GC were incubated in 10% serum, were washed, and were solubilized in the zwitterionic sulfobetaine detergent SB12. Immunoprecipitation of 125I-GC from the extract with anti-C5 Sepharose was followed by 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis autoradiography of immunoprecipitated material. Polyacrylamide gel analysis of surface-labeled 125I-GC showed prominent bands for proteins I and III for both serum-resistant GC strain 6305 and serum-sensitive GC strain 7189. These same bands were visible with similar intensity in the SB12 extracts from presensitized and non-presensitized 6305 and 7189 after serum incubation. For those organisms bearing bactericidal C5b-9 (6305 + IgG and 7189 +/- IgG), additional distinctive bands immunoprecipitated with antibody to C5 Sepharose. These components were of 93,000, 44,000 40,000, and 15,000 daltons for 6305 + IgG, and were of 90,000, 50,000, 44,000, and 19,000 daltons for 7189 +/- IgG. Nonbactericidal C5b-9 extracted from the surface of 6305 incubated in serum, but not sensitized with antibody, was not associated with these distinctive proteins. However, this nonbactericidal C5b-9 did have a different pattern of associated bacterial surface constituents from that observed in control samples incubated with antibody to human serum albumin, which were similar to those with nonserum-incubated organisms. These studies support our earlier experiments which demonstrated that C5b-9 is in a different molecular configuration on the surface of serum-resistant GC from that on the surface of serum-sensitive GC or resistant GC rendered sensitive with bactericidal antibody.