- Email: [email protected]
Melanin and HIV in sub-Saharan Africa
Journal of Theoretical Biology 223 (2003) 131–133
Melanin and HIV in sub-Saharan Africa John T. Manninga,*,1, Peter E. Bundredb, P. Henzic,2 a
Population Biology Research Group, School of Biological Sciences, University of Liverpool, P.O. Box 147, Liverpool L69 3BX, UK b Department of Primary Care, University of Liverpool, Liverpool, L69 3GX, UK c School of Anthropology and Psychology, University of Natal, Durban, South Africa Received 16 September 2002; accepted 13 January 2003
Abstract HIV is common in sub-Saharan Africa. Sexually transmitted bacterial and fungal infections increase the chance of HIV infection. Melanin can prevent the penetration of skin and mucus membranes by microorganisms, and soluble melanin can inhibit HIV replication. We suggest that melanin may reduce the incidence of HIV infection through venereally acquired skin lesions, thus reducing the risk of sero-conversion and slow the progress to AIDS. Indigenous sub-Saharan peoples are highly melanized, but there is pigment variation between populations. We show that skin reflectance, a negative correlate of melanin, is positively associated with adult rate of HIV in sub-Saharan countries. There is no such relationship in populations outside sub-Saharan Africa. We suggest that melanin concentration in black people may correlate with resistance to HIV infection. r 2003 Elsevier Science Ltd. All rights reserved. Keywords: Melanin; HIV susceptibility; Sub-Saharan Africa
1. Introduction At present some 28 million of the 40 million people infected with HIV live in sub-Saharan Africa (UNAIDS, 2000). Sexual behaviour constitutes the main risk factor for HIV transmission. However, susceptibility to infection is an important secondary variable which determines patterns of spread. It is therefore of great interest to identify the factors in sub-Saharan Africa which determine susceptibility to HIV infection and subsequent development of symptoms. One possible protective mechanism is the role of melanin. Melanocytes make the organelle (the melanosome) which contains melanin, and the melanosomes are then exported to other neighbouring cell types (Sturm et al., 2001). There is evidence that melanin may act as photoprotection (Jablonski and Chaplin, 2000). However, this may not be its only, or even its main function (Robbins, 1991). Mackintosh, (2001) has convincingly argued that
*Corresponding author. Tel.: +4401517945026; fax: +4401517945094. E-mail address: [email protected] (J.T. Manning). 1 Present address: Department of Psychology, University of Central Lancashire, Preston PR1 2HE, UK. 2 Present address: Bolton Institute of Higher Education, Bolton BL3 5AB, UK.
in addition to blocking UV light, melanocytes, melanosomes and melanin have anti-microbial properties. Thus melanocytes may process and present antigens to CD4+ T cells in an antigen-specific MHC II restricted manner, they can also act as phagocytes against microorganisms, and the melanin polymer by virtue of its size may act as a barrier against microorganisms. The enzyme cascade which leads to melanin, the prophenoloxidase-activating system, forms the central antimicrobial defense system in many animals. This cascade is activated by cell wall components of bacteria, molds and yeast, and the presence of blood parasites such as Plasmodium. The melanin polymer then traps microorganisms within humoral capsules. Most melanins exist in insoluble form, but synthetic soluble melanins have been shown to inhibit HIV-1 replication and cytotoxicity (Montefiori et al., 1990; Montefiori and Zhou, 1991). People with sexually transmitted diseases such as syphilis and gonorrhoea have an elevated risk of HIV infection (Weiss et al., 2000; Szabo and Short, 2000). We suggest that melanin may protect against HIV by reducing the incidence of infection via venereally acquired sores, which include the lesions of chancroid, syphilis, lymphogranuloma venereum, granuloma inguinale and genital herpes. In support of our suggestion
0022-5193/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0022-5193(03)00070-5
J.T. Manning et al. / Journal of Theoretical Biology 223 (2003) 131–133
132
we present evidence that within sub-Saharan Africa indigenous populations with low melanin scores have high rates of HIV, while outside sub-Saharan Africa there is no association between melanin and rates of HIV.
of HIV and reflectance scores suggesting that countries with high HIV rates have light skins relative to those with low rates (b ¼ 0:54; F ¼ 8:27; r2 ¼ 0:27; p ¼ 0:009; Fig. 1a). There was no relationship between HIV rates and reflectance scores in the remaining countries (b ¼ 0:01; F ¼ 1:74; r2 ¼ 0:03; p ¼ 0:19; Fig. 1b).
2. Melanin and HIV 3. Discussion The concentration of melanin in the skin is measured in reflectance scores. Melanin absorbs light, therefore high scores are associated with low melanin concentrations. Jablonski and Chaplin (2000) list reflectance scores from 82 indigenous populations, of which 25 are from sub-Saharan Africa, and most of the remainder are from other Old World countries. Estimates of % adult rate of HIV for these countries were taken from UNAIDS (2000). Within sub-Saharan Africa we found a significant positive relationship between % adult rate
40
HIV (% Adult Rate)
35 30 25 20 15 10 5 0 15
20
25
30
35
40
45
50
55
65
70
Observed skin reflectance
(a) 4.5
HIV (% Adult Rate)
4 3.5 3 2.5 2 1.5 1 0.5 0 30
(b)
35
40
45
50
55
60
Observed skin reflectance
Fig. 1. The relationship between skin reflectance scores, a negative correlate of melanin concentration, and % adult rate of HIV in (a) sub-Saharan populations and (b) populations outside sub-Saharan Africa. Reflectance scores obtained from Jablonski and Chaplin (2000) and HIV rates from UNAIDS (2000).
The indigenous peoples of sub-Saharan Africa are highly melanized with higher reflectance scores than expected when considering values of annual average UV light (Jablonski and Chaplin, 2000). This may be because pathogen loads are high in this region (Mackintosh, 2001). The rates of HIV infection are also very high in sub-Saharan Africa. Clearly melanization in itself is not sufficient to protect against high rates of HIV, and overall worldwide patterns of infection may be largely influenced by variation in sexual activity. However, we are not concerned with the overall rate of HIV infection, rather we seek to understand the patterns of infection within the sub-Saharan region. Our demonstration of a negative relationship between melanin and HIV infection rates per country in subSaharan Africa is of course an approximation to the real situation. The country-wide rates of HIV infection are relatively coarse measures because HIV prevalence in Africa is in an unstable and dynamic state, and HIV infection rates may vary substantially between cities within countries (Buve et al., 2001). What is needed are direct studies of within-population variation in melanin scores and their relationship with HIV status. If melanin is an important factor in susceptibility to infection by HIV it is obviously not the only such variable. Other important factors are likely to include: (a) Circumcision—it appears that HIV often enters males via the foreskin, and circumcision may therefore be protective against entry of HIV (Caldwell and Caldwell, 1996; Bailey et al., 2001). The foreskin is rich in Langerhans cells that have HIV receptors on their surface. This may be the main point of entry into the penis of an uncircumcised man (Szabo and Short, 2000). We do not envisage that melanin is directly relevant to this process. Rather we suggest that it provides a barrier to bacteria and fungi that cause sores which facilitate entry into the foreskin and into the vulva. The melanin concentration in black foreskins has been shown to average almost three times that found in white foreskins (Iwata et al., 1990). The vulva is also rich in melanin (Jones, 1979), and higher concentrations are likely in black women compared to white. (b) Testosterone— prenatal and adult testosterone are thought to suppress immune function (Geschwind and Galaburda, 1985; Folstad and Karter, 1992). A proxy for prenatal testosterone, the ratio between the length of 2nd and
J.T. Manning et al. / Journal of Theoretical Biology 223 (2003) 131–133
4th digits (2D:4D ratio), has been found to be very masculinized in black South Africans (Manning et al., 2000). This and other evidence has led to the suggestion that high prenatal testosterone concentrations may predispose sub-Saharan populations to high HIV/AIDS rates (Manning et al., 2001). Melanin concentrations are sex dependent in that all known human populations show darker skin in males than females (Jablonski and Chaplin, 2000). The sex difference is likely to be in part dependent on testosterone. Testosterone tends to increase melanin production and estrogen stimulates a reduction in melanin in a variety of species including humans (Wilson and Spaziani, 1976; Wilson 1983; Jee et al., 1994). Thus melanin may have a role as a protective against penetration of skin pathogens in individuals with reduced immune competence. We conclude that individual differences in melanin as measured by reflectance scores may be found to be predictive of susceptibility to HIV infection in the subSaharan region of Africa. Within-population variance in melanin concentration may be a useful public health tool in the identification of individuals with high susceptibility to infection with HIV.
References Bailey, R.C., Plummer, F.A., Moses, S., 2001. Male circumcision and HIV prevention: current knowledge and future research directions. Lancet Infect. Dis. 1 (4), 223–231. Buve, A., Lagarde, E., Carael, M., Rutenberg, N., Ferry, B., Glynn, J.R., Laourou, M., Akam, E., Chege, J., Sukwa, T., 2001. Interpreting sexual behaviour data: validity issues in the multicentre study on factors determining the differential spread of HIV in four African cities. AIDS 15 (Suppl. 4), S117–S126. Caldwell, J., Caldwell, P., 1996. The African Aids epidemic. Sci. Am., 62–69. Folstad, I., Karter, A.J., 1992. Parasites, bright males and the immunocompetence handicap theory. Am. Nat. 139, 603–622. Geschwind, N., Galaburda, A.M., 1985. Cerebral lateralization. Biological mechanisms, associations and pathology: a program for research. Arch. Neurol. 42, 428–459.
133
Iwata, M., Corn, T., Iwata, S., Everrett, M.A., Fuller, B.B., 1990. The relationship between tyrosinase activity and skin color in human foreskins. J. Invest. Dermatol. 95, 9–15. Jablonski, N.G., Chaplin, G., 2000. The evolution of skin coloration. J. Hum. Evol. 39, 57–106. Jee, S.H., Lee, S.Y., Chiu, H.C., Chang, C.C., Chen, T.J., 1994. Effects of estrogen and estrogen receptor in normal human melanocytes. Biochem. Biophys. Res. Commun. 199, 1407–1412. Jones, I.S., 1979. An assessment of vulval pigmentation. N. Z. Med. J. 89, 348–350. Mackintosh, J.A., 2001. The antimicrobial properties of melanocytes, melanosomes and melanin and the evolution of black skin. J. theor. Biol. 211, 101–113. Manning, J.T., Barley, L., Lewis-jones, I., Walton, J., Trivers, R.L., Thornhill, R., Singh, D., Rhode, P., Bereckzei, T., Henzi, P., Soler, M., Sved, M., 2000. The 2nd to 4th digit ratio, sexual dimorphism, population differences and reproductive success: evidence for sexually antagonistic genes. Evol. Hum. Behav. 21, 163–183. Manning, J.T., Henzi, P., Bundred, P.E., 2001. The ratio of 2nd to 4th digit length: a proxy for testosterone, and susceptibility to HIV and AIDS? Med. Hypoth. 57, 761–763. Montefiori, D.C., Zhou, J.Y., 1991. Selective antiviral activity of synthetic soluble l-tyrosine and l-dopa melanins against human immunodeficiency virus in vitro. Antiviral Res. 15, 11–25. Montefiori, D.C., Modliszewski, A., Shaff, D.I., Zhou, J., 1990. Inhibition of human immunodeficiency virus type 1 replication and cytopathicity by synthetic soluble catecholamine melanins in vitro. Biochem. Biophys. Res. Commun. 168, 200–205. Robbins, A.H., 1991. Biological Perspectives on Human Pigmentation. Cambridge University Press, Cambridge. Sturm, R.A., Teasdale, R.D., Box, N.F., 2001. Human pigmentation genes: identification, structure and consequences of polymorphic variation. Genetics 277, 49–62. Szabo, R., Short, R.V., 2000. How does male circumcision protect against HIV infection? Br. Med. J. 320, 1592–1594. Unaids, 2000. Report on the global HIV/AIDS epidemic. Weiss, H.A., Quigley, M.A., Hayes, R.L., 2000. Malecircumcision, risk of HIV infection in sub-Saharan Africa: a systematic review and meta-analysis. AIDS 14, 2361–2370. Wilson, M.J., 1983. Inhibition of development of both androgendependent and androgen-independent pigment cells in scrotal skin dermis of the rat by antiandrogen treatment during fetal growth. Endocrinology 112, 321–325. Wilson, N.J., Spaziani, E., 1976. The melanogenic response to testosterone in scrotal epidermis: effects on tyrosinase activity and protein synthesis. Acta Endocrinol. (Copenh.) 81, 435–448.
Melanin and HIV in sub-Saharan Africa John T. Manninga,*,1, Peter E. Bundredb, P. Henzic,2 a
Population Biology Research Group, School of Biological Sciences, University of Liverpool, P.O. Box 147, Liverpool L69 3BX, UK b Department of Primary Care, University of Liverpool, Liverpool, L69 3GX, UK c School of Anthropology and Psychology, University of Natal, Durban, South Africa Received 16 September 2002; accepted 13 January 2003
Abstract HIV is common in sub-Saharan Africa. Sexually transmitted bacterial and fungal infections increase the chance of HIV infection. Melanin can prevent the penetration of skin and mucus membranes by microorganisms, and soluble melanin can inhibit HIV replication. We suggest that melanin may reduce the incidence of HIV infection through venereally acquired skin lesions, thus reducing the risk of sero-conversion and slow the progress to AIDS. Indigenous sub-Saharan peoples are highly melanized, but there is pigment variation between populations. We show that skin reflectance, a negative correlate of melanin, is positively associated with adult rate of HIV in sub-Saharan countries. There is no such relationship in populations outside sub-Saharan Africa. We suggest that melanin concentration in black people may correlate with resistance to HIV infection. r 2003 Elsevier Science Ltd. All rights reserved. Keywords: Melanin; HIV susceptibility; Sub-Saharan Africa
1. Introduction At present some 28 million of the 40 million people infected with HIV live in sub-Saharan Africa (UNAIDS, 2000). Sexual behaviour constitutes the main risk factor for HIV transmission. However, susceptibility to infection is an important secondary variable which determines patterns of spread. It is therefore of great interest to identify the factors in sub-Saharan Africa which determine susceptibility to HIV infection and subsequent development of symptoms. One possible protective mechanism is the role of melanin. Melanocytes make the organelle (the melanosome) which contains melanin, and the melanosomes are then exported to other neighbouring cell types (Sturm et al., 2001). There is evidence that melanin may act as photoprotection (Jablonski and Chaplin, 2000). However, this may not be its only, or even its main function (Robbins, 1991). Mackintosh, (2001) has convincingly argued that
*Corresponding author. Tel.: +4401517945026; fax: +4401517945094. E-mail address: [email protected] (J.T. Manning). 1 Present address: Department of Psychology, University of Central Lancashire, Preston PR1 2HE, UK. 2 Present address: Bolton Institute of Higher Education, Bolton BL3 5AB, UK.
in addition to blocking UV light, melanocytes, melanosomes and melanin have anti-microbial properties. Thus melanocytes may process and present antigens to CD4+ T cells in an antigen-specific MHC II restricted manner, they can also act as phagocytes against microorganisms, and the melanin polymer by virtue of its size may act as a barrier against microorganisms. The enzyme cascade which leads to melanin, the prophenoloxidase-activating system, forms the central antimicrobial defense system in many animals. This cascade is activated by cell wall components of bacteria, molds and yeast, and the presence of blood parasites such as Plasmodium. The melanin polymer then traps microorganisms within humoral capsules. Most melanins exist in insoluble form, but synthetic soluble melanins have been shown to inhibit HIV-1 replication and cytotoxicity (Montefiori et al., 1990; Montefiori and Zhou, 1991). People with sexually transmitted diseases such as syphilis and gonorrhoea have an elevated risk of HIV infection (Weiss et al., 2000; Szabo and Short, 2000). We suggest that melanin may protect against HIV by reducing the incidence of infection via venereally acquired sores, which include the lesions of chancroid, syphilis, lymphogranuloma venereum, granuloma inguinale and genital herpes. In support of our suggestion
0022-5193/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0022-5193(03)00070-5
J.T. Manning et al. / Journal of Theoretical Biology 223 (2003) 131–133
132
we present evidence that within sub-Saharan Africa indigenous populations with low melanin scores have high rates of HIV, while outside sub-Saharan Africa there is no association between melanin and rates of HIV.
of HIV and reflectance scores suggesting that countries with high HIV rates have light skins relative to those with low rates (b ¼ 0:54; F ¼ 8:27; r2 ¼ 0:27; p ¼ 0:009; Fig. 1a). There was no relationship between HIV rates and reflectance scores in the remaining countries (b ¼ 0:01; F ¼ 1:74; r2 ¼ 0:03; p ¼ 0:19; Fig. 1b).
2. Melanin and HIV 3. Discussion The concentration of melanin in the skin is measured in reflectance scores. Melanin absorbs light, therefore high scores are associated with low melanin concentrations. Jablonski and Chaplin (2000) list reflectance scores from 82 indigenous populations, of which 25 are from sub-Saharan Africa, and most of the remainder are from other Old World countries. Estimates of % adult rate of HIV for these countries were taken from UNAIDS (2000). Within sub-Saharan Africa we found a significant positive relationship between % adult rate
40
HIV (% Adult Rate)
35 30 25 20 15 10 5 0 15
20
25
30
35
40
45
50
55
65
70
Observed skin reflectance
(a) 4.5
HIV (% Adult Rate)
4 3.5 3 2.5 2 1.5 1 0.5 0 30
(b)
35
40
45
50
55
60
Observed skin reflectance
Fig. 1. The relationship between skin reflectance scores, a negative correlate of melanin concentration, and % adult rate of HIV in (a) sub-Saharan populations and (b) populations outside sub-Saharan Africa. Reflectance scores obtained from Jablonski and Chaplin (2000) and HIV rates from UNAIDS (2000).
The indigenous peoples of sub-Saharan Africa are highly melanized with higher reflectance scores than expected when considering values of annual average UV light (Jablonski and Chaplin, 2000). This may be because pathogen loads are high in this region (Mackintosh, 2001). The rates of HIV infection are also very high in sub-Saharan Africa. Clearly melanization in itself is not sufficient to protect against high rates of HIV, and overall worldwide patterns of infection may be largely influenced by variation in sexual activity. However, we are not concerned with the overall rate of HIV infection, rather we seek to understand the patterns of infection within the sub-Saharan region. Our demonstration of a negative relationship between melanin and HIV infection rates per country in subSaharan Africa is of course an approximation to the real situation. The country-wide rates of HIV infection are relatively coarse measures because HIV prevalence in Africa is in an unstable and dynamic state, and HIV infection rates may vary substantially between cities within countries (Buve et al., 2001). What is needed are direct studies of within-population variation in melanin scores and their relationship with HIV status. If melanin is an important factor in susceptibility to infection by HIV it is obviously not the only such variable. Other important factors are likely to include: (a) Circumcision—it appears that HIV often enters males via the foreskin, and circumcision may therefore be protective against entry of HIV (Caldwell and Caldwell, 1996; Bailey et al., 2001). The foreskin is rich in Langerhans cells that have HIV receptors on their surface. This may be the main point of entry into the penis of an uncircumcised man (Szabo and Short, 2000). We do not envisage that melanin is directly relevant to this process. Rather we suggest that it provides a barrier to bacteria and fungi that cause sores which facilitate entry into the foreskin and into the vulva. The melanin concentration in black foreskins has been shown to average almost three times that found in white foreskins (Iwata et al., 1990). The vulva is also rich in melanin (Jones, 1979), and higher concentrations are likely in black women compared to white. (b) Testosterone— prenatal and adult testosterone are thought to suppress immune function (Geschwind and Galaburda, 1985; Folstad and Karter, 1992). A proxy for prenatal testosterone, the ratio between the length of 2nd and
J.T. Manning et al. / Journal of Theoretical Biology 223 (2003) 131–133
4th digits (2D:4D ratio), has been found to be very masculinized in black South Africans (Manning et al., 2000). This and other evidence has led to the suggestion that high prenatal testosterone concentrations may predispose sub-Saharan populations to high HIV/AIDS rates (Manning et al., 2001). Melanin concentrations are sex dependent in that all known human populations show darker skin in males than females (Jablonski and Chaplin, 2000). The sex difference is likely to be in part dependent on testosterone. Testosterone tends to increase melanin production and estrogen stimulates a reduction in melanin in a variety of species including humans (Wilson and Spaziani, 1976; Wilson 1983; Jee et al., 1994). Thus melanin may have a role as a protective against penetration of skin pathogens in individuals with reduced immune competence. We conclude that individual differences in melanin as measured by reflectance scores may be found to be predictive of susceptibility to HIV infection in the subSaharan region of Africa. Within-population variance in melanin concentration may be a useful public health tool in the identification of individuals with high susceptibility to infection with HIV.
References Bailey, R.C., Plummer, F.A., Moses, S., 2001. Male circumcision and HIV prevention: current knowledge and future research directions. Lancet Infect. Dis. 1 (4), 223–231. Buve, A., Lagarde, E., Carael, M., Rutenberg, N., Ferry, B., Glynn, J.R., Laourou, M., Akam, E., Chege, J., Sukwa, T., 2001. Interpreting sexual behaviour data: validity issues in the multicentre study on factors determining the differential spread of HIV in four African cities. AIDS 15 (Suppl. 4), S117–S126. Caldwell, J., Caldwell, P., 1996. The African Aids epidemic. Sci. Am., 62–69. Folstad, I., Karter, A.J., 1992. Parasites, bright males and the immunocompetence handicap theory. Am. Nat. 139, 603–622. Geschwind, N., Galaburda, A.M., 1985. Cerebral lateralization. Biological mechanisms, associations and pathology: a program for research. Arch. Neurol. 42, 428–459.
133
Iwata, M., Corn, T., Iwata, S., Everrett, M.A., Fuller, B.B., 1990. The relationship between tyrosinase activity and skin color in human foreskins. J. Invest. Dermatol. 95, 9–15. Jablonski, N.G., Chaplin, G., 2000. The evolution of skin coloration. J. Hum. Evol. 39, 57–106. Jee, S.H., Lee, S.Y., Chiu, H.C., Chang, C.C., Chen, T.J., 1994. Effects of estrogen and estrogen receptor in normal human melanocytes. Biochem. Biophys. Res. Commun. 199, 1407–1412. Jones, I.S., 1979. An assessment of vulval pigmentation. N. Z. Med. J. 89, 348–350. Mackintosh, J.A., 2001. The antimicrobial properties of melanocytes, melanosomes and melanin and the evolution of black skin. J. theor. Biol. 211, 101–113. Manning, J.T., Barley, L., Lewis-jones, I., Walton, J., Trivers, R.L., Thornhill, R., Singh, D., Rhode, P., Bereckzei, T., Henzi, P., Soler, M., Sved, M., 2000. The 2nd to 4th digit ratio, sexual dimorphism, population differences and reproductive success: evidence for sexually antagonistic genes. Evol. Hum. Behav. 21, 163–183. Manning, J.T., Henzi, P., Bundred, P.E., 2001. The ratio of 2nd to 4th digit length: a proxy for testosterone, and susceptibility to HIV and AIDS? Med. Hypoth. 57, 761–763. Montefiori, D.C., Zhou, J.Y., 1991. Selective antiviral activity of synthetic soluble l-tyrosine and l-dopa melanins against human immunodeficiency virus in vitro. Antiviral Res. 15, 11–25. Montefiori, D.C., Modliszewski, A., Shaff, D.I., Zhou, J., 1990. Inhibition of human immunodeficiency virus type 1 replication and cytopathicity by synthetic soluble catecholamine melanins in vitro. Biochem. Biophys. Res. Commun. 168, 200–205. Robbins, A.H., 1991. Biological Perspectives on Human Pigmentation. Cambridge University Press, Cambridge. Sturm, R.A., Teasdale, R.D., Box, N.F., 2001. Human pigmentation genes: identification, structure and consequences of polymorphic variation. Genetics 277, 49–62. Szabo, R., Short, R.V., 2000. How does male circumcision protect against HIV infection? Br. Med. J. 320, 1592–1594. Unaids, 2000. Report on the global HIV/AIDS epidemic. Weiss, H.A., Quigley, M.A., Hayes, R.L., 2000. Malecircumcision, risk of HIV infection in sub-Saharan Africa: a systematic review and meta-analysis. AIDS 14, 2361–2370. Wilson, M.J., 1983. Inhibition of development of both androgendependent and androgen-independent pigment cells in scrotal skin dermis of the rat by antiandrogen treatment during fetal growth. Endocrinology 112, 321–325. Wilson, N.J., Spaziani, E., 1976. The melanogenic response to testosterone in scrotal epidermis: effects on tyrosinase activity and protein synthesis. Acta Endocrinol. (Copenh.) 81, 435–448.