Z. Morph. Anthrop.

81

365-371

Department

Stuttgart,November

1997

of Human

Genetics, and Department

of Anthropology Department Andhra University,

Visakhapatnam,

of Statistics, India

Genetic structure of Rajaka caste and affinitieswith other caste populations of Andhra Pradesh, India By C. Parvatheesam,

B. V. Babu

and M. C. Babu

With 1 figureand 2 tables in the text Summary:

The

present

study gives

an account

of the genetic

structure

in terms of distribu

tion of a few geneticmarkers, viz., A1A2B0, Rh(D), G6PD deficiencyand haemoglobin the Rajaka

caste

of Andhra

India. The genetic relationships of the Pradesh, were in terms of genetic investigated populations esta and Sanghvi. Relatively lesser distance was distance, i.e., SqB (mn) of Balakrishnan castes. Also, blished between the Rajaka and two Panchama the pattern of genetic distance varna system. corroborates the hierarchical order of the Hindu among

caste with

Rajaka

population other Andhra

caste

Zusammenfassung: Am Beispiel einigergeneticmarkers (A1A2B0, Rh(D), G6PD-Mangel undHamoglobin) wird einUberblick iiberdie genetische MerkmalsverteilunginderRajaka in Andhra

Kaste mit

anderen

Distanzen

der Rajaka-Kaste Pradesh, Indien, gegeben. Die genetischen Beziehungen in Andhra mit Hilfe Pradesh werden der genetischen Kastenbevolkerungen untersucht. Die zwei Unterschiede zwischen den und festgestellten Rajaka

Panchama-Kasten muster

erwiesen

der hierarchischen

sich als relativ gering. Auch hier zeigte sich, dafi die Distanz des Hindu-Varna-Systems folgen.

Ordnung

Introduction The people of India are estimated to consist of over 40,000 Mendelian populations. As estimated 37,000 endogamous groups are structured in a system commonly referred to as theHindu caste system.There are some 3000 Mendelian populations that are, strictlyspeaking, outside the caste system.These populations include tribal autochthonous groups and religious communities (Malhotra 1984). Andhra one 200 of the harbours about Hindu castes, thirty Pradesh, twenty-five states, three tribes and a few religious minority communities. The Hindu castes are ca tegorized according to classical varna system,where the division is primarily into four varnas (classes), namely, Brahmin, Kshatriya, Vyshya and Sudra. Traditionally, the first three varnas consist of priest, ruler or warrior and merchant castes, re a castes such spectively ,while the fourthvarna, Sudra, consists of large number of as peasants, artisans, servicing castes, etc. The fifthclass, Panchama, which is out side the caste system, includes all castes which were once considered as un touchable.

Various studies have been undertaken on the distribution of genetic markers among tribalpopulations, which have revealed a considerable genetic heterogeneity within and among these populations. However, only a few studies are available on

0044-314X/97/0081-0365 ?

1997 E. Schweizerbart'sche Verlagsbuchhandlung, D-70176

$ 1.75 Stuttgart

366

C.

Parvatheesam,

B. V. Babu

and M. C.

Babu

caste populations. Hence, the present study aimed to report the genetic structureof theRajaka caste ofAndhra Pradesh. The Rajaka, awashermen caste of Sudra varna, is distributed all over the State ofAndhra Pradesh. The Rajaka is also referred to as Chakali or Dhobi. In fact, theword Dhobi is used tomean washerman by Anglo Indians (Thurston 1975). A majority of Rajaka are engaged inwashing clothes of other communities and, of late, a few are in government and private employment. Among the Rajaka many patrilineal exogamous clans exist, inwhich consangui neous

marriages

are

preferred.

They

are

non-vegetarians

but without

beef

con

sumption, and speak Telugu, a language of theDravidian language family. This study also aimed to analyze the genetic variation in terms of genetic distance among some Andhra castes belonging to four varnas in order to examine the affinitiesof Rajaka with other castes. The genetic distance analysis is a simple method to investigate the relationship among a set of populations. Also, thepattern of relationships can be derived to get some idea about the forces which have operated in the evolutionary history of these populations. It is endeavoured to consider eleven castes to represent all varnas and aMuslim religious group. They include Brahmin, Vyshya (representing Brahmin and Vyshya varnas, respectively), Kapu, Kamma, Reddi (peasant castes of Sudra varna), Padmasali, Rajaka (artisan and servicing castes of Sudra varna, respectively),Mala, Madiga, Relli (Panchama varna) and Muslim, a religious minority group. Thus, the objectives of the present study are (1) to analyze the distribution of a few genetic markers, viz., A1A2B0 and Rh(D) blood groups, glucose-6-phosphate dehydrogenase (G6PD) and haemoglobin among theRajaka caste and (2) to study the extent of genetic relationships between theRajaka and other caste populations of Andhra Pradesh, India. Materials

and methods

investigations on genetic markers, viz., A1A2B0 and Rh(D) blood groups, glucose-6-phosphate dehydrogenase (G6PD) and haemoglobin were carried out among unrelated Rajaka individuals inhabitingVisakhapatnam city,Andhra Pra desh. Blood samples were collected intoNa EDTA. Blood cells and serum were separated and blood group typingwas performed by using appropriate antisera. Haemolysates were prepared and stored at -20 ?C. The electrophoresis forG6PD and haemoglobin was performed on cellulose acetate Titan III plates (Helena La 8c Alperin boratories, Texas), according to Ellis (1972). The gene frequency calculations were calculated by themaximum likelihood method as described by Balakrishnan (1988 a). The details of two other markers (PTC taste sensitivity and colour blindness) among theRajaka, published earlier by us (Parvatheesam & Babu 1996), were also utilised in the present study. The distribution of all these geneticmarkers (Al A2B0, Rh(D), G6PD, Hb, PTC and CB) among ten other caste 1994, Bhas populations from published and unpublished works (Babu & Naidu kar & Reddy 1977,Chandraiah & Bahadur 1979,Gopalam 1980, Khaja 1993, Lakshmi 1986,Naidu et al. 1978, 1990; Ramaswamy 1984, Ramesh 1981,Naidu 1992, Ranganayaki & Injeti 1979, Reddy 1985, Reddy et al. 1980, 1982 and Srikumari 1985) was considered for genetic distance analysis. Only ten popula tionswere obtained for the distribution of all these sixmarkers. All varnas except Kshatriya along with theMuslim religious group were represented for the present analysis. No systematic genetic studies were available on theKshatriya. The

Genetic

structure of Rajaka

367

caste and affinities

The genetic distance analysis was carried out by index B of Balakrishnan & Sanghvi (1968), since it is designed to discriminate among various populations with amaximum efficiency in a statistical sense. This value considers the correlation among the variables and states each character of genetic distance derived directly from the disperson matrix obtained during themaximum likelihood estimation of gene frequencies. The relative distance values obtained by this value were standar dised proportionately so as tomake the largest squared distance equal to 99. The squared distances were formed into clusters by using the pair group method. Based on thematrix obtained from cluster analysis a dendrogram was constructed. All these analyses were performed according tomethods illustrated by Balakrishnan (1988 b,c).

Results

and discussion

The phenotype and allele frequencies of four geneticmarkers, i. e.,Al A2B0, Rh(D), G6PD and haemoglobin are shown inTable 1.Among theRajaka, blood group B marked the highest incidence followed by the 0 group. However, all in all the allele 0 marked the highest frequency. The D/a value (0.3533) and x2 value (0.1698) computed for genetic equilibrium indicate that the Rajaka is inHardy-Weinberg equilibrium for this locus. Reviewing the ABO phenotype frequencies of other Andhra caste populations (Parvatheesam 1995), it can be observed that a majority of studies (29 out of 38) recorded higher frequencies of blood group 0 than eitherB or A, while a few (9 out of 38) reported the highest frequencies of group B. Compared to the average frequency of the A gene in Andhra castes, the tribes recorded higher frequencies, which confirms that Andhra castes have a higher frequency of the B gene thanAndhra tribes (Babu 1993). The incidence of Rh(D) negative is 6.77% with a d allele frequency of 0.2601. This is generally in linewith

1. Phenotype

Table

and allele

Pradesh.

Locus

frequencies

Phenotype

A1A2B0 = (n 133) A2B

of genetic markers

Allele Total observed No.

22 Al 16.54 A2 4.51 6 A2 B 53 B 39.85 8 A1B 6.02 2.26 3 41 30.83 0 D 93.23 6.77

G6PD

57 G+ G-3

95.00 5.00 GdB"

Hb

100 100.00 AA 0.00 0 AS 0.00 0 SS

(n

100)

frequency

0

+ve 124 9 -ve

=

%

Al

Rh(D) = (n 133) = Only men (n 60)

caste of Andhra

among Rajaka

d

=0.1203 = 0.0394 = 0.2805 = 0.5598

= 0.7399 = 0.2601 GdB+ = 0.9786 = 0.0214 HbA = 1.0000 Hbs = 0.0000

368

C. 2. Genetic

Table

B. V. Babu

Parvatheesam,

distance

and

(Sq B)

and M.

C.

Babu for 11 caste populations

F values

corresponding

Andhra Pradesh (Based on six polymorphic loci).

(1) (1) Brahmin (2) Vyshya (3) Kapu (4) Kamma (5) Reddi (6) Padmasali (7) Rajaka (8)Mala (9)Madiga (10) Relli (11)Muslim Figures Figures * Non

(2)

(3)

(4)

(5)

(6)

(7)

(8)

7.9

10.9

6.3

30.8

14.3

79.1

28.6

10.4

87.7

41.5

105.0

58.7

70.2

177.0

172.0

50.0

342.0

22.2

41.8

96.2

61.5

10.4

142.7

24.4

22.9

59.8

23.4

8.6

135.7

51.5

167.5

85.6

9.7

204.0

59.6

95.3

74.8

17.2

183.3

11.3

11.7

30.3

31.6

17.9

118.0

174.8

2.32 2.18

2.34

1.4*

34.4

18.3

0.38

4.51

0.08

0.89

2.43

3.58

1.70

1.63

0.26

1.41

5.71

4.04

2.13

3.92

5.75

4.19

6.37

5.46

2.49

5.36

4.78

1.66

4.22

3.47

1.09

3.17

4.13

0.88

0.54

0.60

2.05

0.59

0.40

0.36

0.95

2.28

3.58

2.10

8.14

6.41

7.64 10.34

2.42

5.32

4.58

5.24 14.15

in the upper

triangle triangle

(10) (11)

(9)

37.1

53.5

3.44 0.63 0.76

in the lower

of

are are

significant; Remaining

genetic

4.5

distance

0.81

at 0.1 %

201.0

7.14

values.

[sq B(m,n)]

F(7>a) values. values are significant

129.0

56.9

level.

those observed in other Andhra caste populations. The tribal populations of this region differwith castes. Several tribes in this region exhibited monomorphism for Rh D (Babu 1993). The X-linked G6PD deficiency is noticed among 5% of Rajaka men. Women lack this phenotype. The frequency of GdB" is 0.0214. The G6PD deficiency cases so farobserved inAndhra Pradesh, resemble theMediterranean B type.However, fast and slow variants (G6PD AB and CB) were reported among the Jalari caste of this region (Naidu et al. 1985). The Rajaka individuals in this study show only adult haemoglobin. The sickle cell gene was not encountered.

Rajaka Madiga Mala

Muslim Brahmin

Relli Kapu

Reddi

r-P-?

Padmasali Kamma

Vysya Fig.

1.

Dendrogram

for genetic

relationships

among

11 Andhra

caste

populations.

Genetic

structure of Rajaka

369

caste and affinities

Among Andhra caste populations, the sickle cell gene was found among a few, namely, Jalari (Naidu et al. 1985), Kapu, Mala, Madiga (Ramaswamy 1984) and Relli (Ramesh 1992). The genetic relationships of theRajaka with ten otherAndhra caste populations have been estimated and presented in Table 2. The Rajaka exhibits significantly higher genetic distance to castes of upper varnas (Brahmin and Vyshya) and upper Sudra castes (Kapu, Kamma, Reddi and Padmasali). The distance is relatively lower between the Rajaka and two castes of Panchama varna, viz., Mala and Madiga. Fig. 1 displays the dendrogram which was constructed on the basis of the distance matrix. It is evident from the figure that all eleven castes have fallen into two main clusters. The first cluster includes Rajaka, Mala, Madiga and Muslim. Mala and Madiga, which are allied castes of panchama varna, inhabit together, formed into one cluster. The closest relationship of Rajaka to this subcluster is also under standable as Rajaka falls in the last rung of Sudra hierarchy afterwhich only Panchama varna exists towhich theMala and Madiga castes belong. The presence of theMuslim group in this cluster could be explained on the ethnohistorical basis that a large proportion of the gene pool in present day Muslims of this region is derived from the converts particularly from low Hindu castes during theMoghul rule in themedieval period. At that time low casteHindus were facing humiliation under the tyranny of the caste systems.However, this conversion must have been either by force, by offer ofmaterial enrichment or by pressure byMuslim invaders (Jane 1985). In the second cluster,most intriguing is the formation of a cluster between Brahmin and Relli. Brahmin is at the top of the hierarchy,whereas Relli, a Panchama caste, are said to be migrants from the neighbouring State ofOrissa. It is interesting to note thatBrahmin and Relli castes have shared a common MtDNA haplotype, and that an unrooted neighbour joining network, linking caste popula tions, also joins both these castes, as was shown in a study on mitochondrial DNA variation among Andhra castes (Bamshad et al. 1996). This supports the possibility et al. of marriages between high and low castes as suggested by Chakraborty seen between is between Sudra Close castes, (1986). upper relationship particularly the allied castes, Kapu and Reddi, which indicates that theymight have diverged from

a common

ancestral

group.

Vyshya

is distant

to social

rank. The

from

these

groups.

Thus,

the

genetic distance analysis of eleven caste populations in this study reveals a relatively higher degree of corroboration with the hierarchical system of theHindu varna system.However, Balakrishnan (1978) reported theminimally structured genetic variation

among

castes

according

varna

system

is characterised

by substantial geographical structuringof populations and the possibility of gene flow and admixture between populations of some geographical regions which can be predicted from an ethnohistorical basis. The pattern of genetic variation among these populations are complex and hence inferencesmade from these studies should be viewed cautiously, since these studies have only considered a few genetic loci, and gene frequency changes are subject to large stochastic errors (Nei 1987). References Babu, B. V. (1993):A population geneticstudyof sub tribesofMali fromAndhra Pradesh. Ph. D.

thesis, Andhra

University,

Visakhapatnam.

Babu, B. V. & Naidu, J.M. (1994):A1A2B0 andRh(D) blood groups among someAndhra populations.

-

J.Hum.

Eco.

5, 301-302.

C.

370

B. V. Babu

Parvatheesam,

and M. C.

Babu

some V. (1978): A preliminary Balakrishnan, populations study of genetic distances among 67-75. Hum. Evol. of the Indian subcontinent. 7, J. - In: Malho and allele frequency estimation. (1988a): Hardy-Weinberg equilibrium ISI & Genetics. in Human K. C. Statistical Methods tra, IBRAD, (ed.): Population ISHG, Calcutta, pp. 39-93. K. C. (ed.): In: Malhotra, distance: definition and computation. (1988b): Genetic & ISI inHuman Genetics. Statistical Methods ISHG, Calcutta, IBRAD, Population pp.

94-139.

In: Malhotra, K.C. (ed.): Statistical Methods analysis. ISI & ISHG, Calcutta, Genetics. IBRAD, pp. 140-207. Population on the basis L. D. V. & Sanghvi, between populations Balakrishnan, (1968): Distance Biometrics attributed data. 24, 859-865. -

(1988c): Human

Genetic

in

distance:

R. L., Busi, B. R., Naidu, A. E., Crawford, M. H., Cann, M., Fraley, in caste populations of Andhra variation Pradesh, (1996): mtDNA Jord, L.B. Hum. Biol. 68, 1-28.

Bamshad,

of

J.M. & India.

Bhaskar, S. &; Reddy, V. R. (1977): The ABO blood groups in Padmasalis of Cuddapah

Ind. Anthrop. 2, 150-153. et al. (1986): Gene differentiation B.N. H., Mukherjee, among Amer. J. Phys. Anthrop. India. 71, 295-309. groups ofWest Bengal, endogamous B. (1979): PTC M. & Bahadur, Chandraiah, groups of tasting gene among six endogamous Ind. J. Phys. Anthrop. Hum. Genet. Gadwal 51, 179-182. (A.P.). for electrophoresis of erythrocyte glucose Ellis, N. & Alperin, J.B. (1972): A rapid method on cellulose acetate Amer. J. Clin. Path. 57, 534 plates. 6-phosphate dehydrogenase 536. district, Andhra

Chakraborty, ten

Pradesh.

R., Walter,

K. B. (1980): Gopalam, Ph. D. Pradesh.

Distribution

of some

genetic markers

thesis, Osmania University, P. T. Genetic studies on some Muslim Jane, (1985): Ph. D. thesis, Osmania University, Hyderabad.

Khaja,

N. MD.

(1993):

Genetic

variations

among

Hyderabad. endogamous

inMuslims.

-

Ph.D.

Vysyas

of Andhra

sects of Andhra

Pradesh.

thesis, Andhra

University,

-

Visakhapatnam. castes of Andhra Pradesh. K.P. (1986): Genetic Hindu studies on some endogamous Lakshmi, Ph. D. thesis, Osmania University, Hyderabad. structure among caste cluster of Maharashtra, K.C. Malhotra, (1984): Population Dhangar In: Lukacs, India. J.R. (ed.): The People of South Asia, The Biological Anthropology of India, Pakistan and Nepal. Plenum Press, New York, pp. 295-324. of Naidu, J.M. (1981): A sero-biochemlcal genetic study of Brahmin and Kamma populations Ph. D. thesis, Andhra University, Visakhapatnam. Coastal Andhra Pradesh. H.W & Neel, Naidu, J.M., Mohrenweiser, J.V. (1985): A sero-biochemical genetic study caste of Jalari and Brahmin India. of Andhra Hum. Hered. Pradesh, 35, populations 148-156. Naidu,

P. & Mukherjee, J.M., Veerraju, in Reddis secretion and Kammas Phys. Anthrop. J.M., Yasmin

D. P. (1978): Blood and ABH saliva groups of Southern Andhra Amer. India. Pradesh, J.

48, 533-543. Sachi Devi,

S. (1990): Further data on ABO and Rh blood groups Hum. Ecol. 1, 53-54. among Andhra J. populations. Columbia Nei, M. Press, New York. (1987): Molecular genetics. Evolutionary University caste A C. treatise of Parvatheesam, (1995): group of Andhra population Rajaka genetic Pradesh. Ph. D. thesis, Andhra University, Visakhapatnam. taste C. & Babu, B. V. (1996): PTC in Rajaka Parvatheesam, sensitivity and colour blindness caste group of Andhra Pradesh. Ecol. 7, 63-64. J.Hum.

Naidu,

Ramaswamy, enzyme Osmania

B.

&

(1984):

Biochemical

polymorphism University,

and

in human Hyderabad.

genetic populations

studies

on

serum

of Andhra

and proteins Pradesh. Ph.D.

red cell thesis,

structure of Rajaka

Genetic M.

Ramesh,

(1992):

A

genetic

study

on Rellis.

caste and affinities -

Ph.D.

371

thesis, Andhra

University,

Visa

khapatnam. L. & Ranganayaki,

Reddy,

and colour blindness Injeti, M.S. among three (1979): Incidence of PTC caste Ind. J. Phys. Anthrop. Hum. Gen. of Visakhapatnam. endogamous populations 5, 169. A. P., Mukherjee, T. (1980): Distribution B.N. & Ramachandraih, of ABO blood Hum. Hered. 30, 331? groups among four endogamous groups of Andhra Pradesh. 332.

Reddy,

A. P., Mukherjee, (1982): A serological

Homo

B. N., Malhotra, and biochemical

33, 174-182. (1985): Genetic

K. C, genetic

Das,

S. K.

study among

8t Ramachandraiah, T. the Coastal and plateau.

K.R. of Cuddapah and genetic variation of theMadigas demography M. Phil, dissertation, S.V. University, Tirupati. Pradesh. district, Andhra C. R. (1985): A study of natural selection and genetic adaptation Srikumari, among four D. Pradesh. Ph. Andhra sub of Andhra thesis, University, populations endogamous

Reddy,

Visakhapatnam. E. (1975): Thurston, Delhi. Address

Caste

and Tribes

of Southern

India

(Reprint).

-

Cosmo

Publications,

for correspondence: Dr.

B. V. Babu, Centre Senior Research Officer, Regional Medical Research 751 016, India. C.S. Pur, Bhubaneswarof Medical Research),

Council

(Indian