On the Genetics of Mandibular Prognathism: Analysis of Large European Noble Families J Med Genet: First Published As 10.1136/Jmg.30.2.112 on 1 February 1993

11217 Med Genet 1993; 30: 112-116 On the genetics of mandibular prognathism: analysis of large European noble families J Med Genet: first published as 10.1136/jmg.30.2.112 on 1 February 1993. Downloaded from

G Wolff, T F Wienker, H Sander

Abstract 21 (81 %) for monozygotic and 2/15 (13%) for Mandibular prognathism is assumed to dizygotic twin pairs,'5 a result which, accord- be a polygenic trait in the vast majority ing to Penrose,'6 strongly argues against a of cases. In a few families, this phenotype monogenic aetiology in most of the affected. and perhaps a syndrome with a broader Taken together, these findings show that in the spectrum of facial anomalies seems to be vast majority of families mandibular prognath- determined by a single dominant gene ism seems to have a polygenic or multifactorial of very low frequency (McKusick No cause. *176700). The phenotype is known to have Nevertheless there are reports of striking occurred independently in several Euro- examples of apparently autosomal dominant pean noble families. We constructed a inheritance of mandibular prognathism, the pedigree comprising 13 of these families best known of which is the Habsburg family with 409 members in 23 generations in together with other European royal fami- which mandibular prognathism has been lies.'7'8 Consanguinity was common in these segregating. Obviously, the presumed families and has previously been suggested as dominant gene is not fully penetrant in accounting for the dominant 'Habsburg jaw"9 the heterozygous state. Pedigree analysis or discussed as a contributing factor acting on using the Elston-Stewart algorithm a multifactorial background.'8 Although it may yields a maximum likelihood estimate be true that consanguinity contributed to some (MLE) of p = 0 955 (SE 0-038) of the pene- of the other disorders which the Habsburg trance parameter. family suffered from, it seems unlikely to be (J Med Genet 1993;30:112-16) responsible for a phenomenon like pseudodo- minance of prognathism.20 There are only a few other reports of familial Mandibular prognathism (McKusick No mandibular prognathism inherited in a way *176700) is one of the best known examples of that suggests the existence of an autosomal a facial genetic trait in humans. According to dominant gene responsible for this phenotype. the classification of Angle,' this phenotype Stiles and Luke'° reported involvement in four http://jmg.bmj.com/ corresponds to class III skeletal malocclusion, generations with apparent non-penetrance in the frequency of which in children was found an obligate gene carrier, and McKusick2' to be in the range of 0 5%2 to 2 7%,3 and in observed the trait in a black family. Thompson mixed and permanent dentition in the range of and Winter20 described a three generation 2 to 4% with a slight preponderance of affected family with mandibular prognathism and males.45 A wide range of environmental factors other facial characteristics similar to those has been suggested as contributing to the de- observed in members of the Habsburg family, velopment of mandibular prognathism,6 but such as thickened lower lip, prominent nose, on September 28, 2021 by guest. Protected copyright. the observation of familial aggregation lends flat malar areas, and mildly everted lower eye- support to the hypothesis that heredity plays a lids. One child had an oxycephalic head shape substantial role in the aetiology. Numerous owing to synostosis of all cranial sutures, a studies have shown a significantly higher inci- feature which the authors suspected also in dence of this phenotype in the relatives of Charles V, a severely affected member of the affected probands.-'0 In the offspring of affec- Habsburg family. ted parents, extensive studies of Japanese It can therefore be hypothesised that there families showed a frequency of 18% if the might be an autosomal dominant gene of mother was affected, 31 % if the father was reduced penetrance possibly leading to a syn- affected, and 40% if both parents were affec- drome of variable clinical expressivity with ted.7"1 In sibs of affected probands, Litton et mandibular prognathism as the main feature Institut fur a16 found a frequency of 13% irrespective of and a variable spectrum of other cranial and Humangenetik und sex. The genetic mechanisms which have been facial anomalies as optional symptoms. Anthropologie, To prove this hypothesis we analysed an University of suggested as being responsible for the pheno- Freiburg i Br, menon of familial aggregation of mandibular extended pedigree of European noble families Breisacher Strasse 33, prognathism include 'irregular' inheritance in which mandibular prognathism was segre- D-7800 Freiburg, gating. Germany. with a penetrance of 70% and variable expres- G Wolff sivity,9 autosomal recessive inheritance,7 12 T F Wienker autosomal dominant inheritance,813 dominant H Sander inheritance as a rule with some exceptions,'4 Methods and results Correspondence to dominant inheritance with incomplete pene- A pedigree was constructed comprising 13 Dr Wolff. trance,'0 and a polygenic threshold model.6 European noble families with 409 members in Received 21 July 1992. for among twin 23 (fig 1). The physical status with Accepted 18 September Concordance prognathism generations 1992. pairs collected from published reports was 17/ regard to mandibular prognathism of each of On the genetics of mandibular prognathism: analysis of large European noble families 113 J Med Genet: first published as 10.1136/jmg.30.2.112 on 1 February 1993. Downloaded from

-U~~~~~~~~~~V

~ ~ ~ ~ ~ i;

A~~~~~~ A~~~

-4A~~~~~

A 0~~~~~~~~~~~~~~A

K -ft~~~~~~~~~~~~~~~"-

A'~~~~~~~~~~~~~i

z* II -4*~~~~~~~~~~~~~

ccA 'A l

;_PXF~~~~~~~~~~~~~~-vP ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~!

-~~~~~~~~~~~'A''-~~~~~~~~~~~~~~~~~~~~~~~t http://jmg.bmj.com/

_7 _

= A

-1CI on September 28, 2021 by guest. Protected copyright.

A~~ ~ ~ ~ ~A~A ~ ~ ;~~~L~~ ~ J IC-m~ -U 14 Wolff, Wienker, Sander

these family members was documented by means of pictures (fig 2) or authentic descrip- 22

tions.2 Family members exhibiting only minor J Med Genet: first published as 10.1136/jmg.30.2.112 on 1 February 1993. Downloaded from symptoms of the trait like everted lower lip were also regarded as affected. Segregation of the phenotype in the offspring of consanguin-

Figure 2C Louis XIV ( 11th generation, fig 1). Wax impression, reproduced from Rene de la Croix Herzog von Castries: Die Bourbonen in Frankreich. In: Die gro,Ben Dynastien. Munchen: Sudwest Verlag, 1978: 170.

Figure 2A Karl V (6th generation, fig 1). Painting of Tizian, reproduced from B Hamann, ed. Die Habsburger. Miinchen: Piper Verlag, 1988: 201 (detail). http://jmg.bmj.com/ on September 28, 2021 by guest. Protected copyright.

Figure 2B Lorenzo il Magnifico di Medici (4th generation, fig 1). Painting of Christofano Figure 2D George II (14th generation, fig 1). dell'Altissimo, reproduced from Karla Langedijk. The Mezzotint engraving of Richard Houston, reproduced portraits of the Medici, 15th-18th centuries. Firenze: from M v Boehn. England im 18. Jtahrhundert. Berlin: Studio per Edizioni Scalte, 1983: 1138. Askanischer Verlag, 1920: 64. On the genetics of mandibular prognathism: analysis of large European noble families 115

but only in the offspring of consanguineous parents who were both affected did the ratio

differ significantly from the 1:1 expected under J Med Genet: first published as 10.1136/jmg.30.2.112 on 1 February 1993. Downloaded from the assumption of autosomal dominant inheritance. Obviously, the gene is not fully penetrant in the heterozygous state. A direct estimate by counting the fraction of 'skipped generations' (fig 4) yielded a crude estimate ofpenetrance of 0 88. In order to obtain an empirical estimate of the penetrance parameter (p), the likelihood of the pedigree was repeatedly calculated using the Elston-Stewart algorithm23 and maximised over this parameter with all other parameters held constant (gene frequency 0-0001, com- plete penetrance in the homozygous state). The LIPED computer program24 was cho- sen since it is a fairly fast and reliable way to handle large and complex pedigrees with mul- tiple consanguinity loops. However, the large pedigree had to be broken down into four parts which did not overlap except for the key ances- tors, otherwise computation times (mainframe IBM 3090/180E, VS-Fortran 2.4.0) exceed all bounds. Thus, we obtained a maximum likeli- III I i. t hood estimate (MLE) of the penetrance parameter, Figure 2E Rudolph I von Habsburg (not included in p (fig 5). the pedigree, the great great grandfather of Ernst der According to standard likelihood theory, the Eiserne (generation 1)). Photograph taken from the variance of the MLE equals the negative in- gravestone in Speyer Cathedral, privately owned by one verse of the second derivative of the likelihood of the authors. function at its maximum. A second order poly- nomial approximation around the maximum25 yields p = 0 955 (SE 0-038). Approximation by eous and non-consanguineous affected parents higher order polynomials (3, 4, 5) did not was analysed by separately estimating the substantially alter the MLE or its standard segregation ratio of affected/unaffected chil- error. Changing the allele frequency up to dren in families with one and both parents 0 001 also did not lead to substantially altered affected (figs 3A-D). In all situations affected results.

offspring were more frequent than unaffected, http://jmg.bmj.com/

Discussion A B Segregation of the mandibular prognathism phenotype in this large pedigree (fig 1) strongly argues in favour of a single dominant gene. The slight preponderance of affected

children among the offspring of affected par- on September 28, 2021 by guest. Protected copyright. ents (fig 3) could be explained by some of the parents being homozygous for the presumed dominant gene. Looking for a possible effect of this suspected homozygous state, we found no 50 44 65 hint of lethality or a more severe clinical expression of the phenotype among the off- C D spring of two affected parents. In concordance with this finding is the observation that most of the very severely affected members, including qw Charles V, did not have parents who were both affected. Generally, ascertainment is a major problem in estimating the penetrance parameter(s) by

93 10 16 3 Figure 3 Segregation of the mandibular prognathism phenotype in the offspring of consanguineous and non-consanguineous affected parents. (A) one parent Figure 4 Estimation of affected + consanguinity, (B) one parent affected, no 121 17 the penetrance of the consanguinity, (C) both parents affected + presumed autosomal consanguinity, (D) both parents affected, no dominant gene by the consanguinity. In all situations affected offspring are direct method (87 7%). more frequent than unaffected, but only in (C) does this observation differ significantly from the ratio expected under the assumption of autosomal dominant inheritance. 116 Wolff, Wienker, Sander Penetrance with malocclusion and functional impair- 0*7 0-8 0.9 1 -° ment2 (fig 2). The clinical spectrum of this

phenotype with regard to more severe facial J Med Genet: first published as 10.1136/jmg.30.2.112 on 1 February 1993. Downloaded from . 170 and perhaps cranial malformations has yet to be determined by careful documentation of other families. 172 0 0 The secretarial assistance of Mrs Wurich is 174 - acknowledged. This paper is dedicated to Pro- fessor Ulrich Wolf on the occasion of his 60th 176 ° birthday on 2 January 1993. 0) 0 178 1 Angle EH. Treatment of malocclusion of the teeth. 7th ed. Philadelphia: White SS Manufacturing Co, 1907:52- 4,58,550-3. 2 Newman GB. Prevalence of malocclusion in children six to 180 fourteen years of age and treatment in preventable cases. 7 Am Dent Assoc 1956;52:566. 3 Seipel CM. Variations of tooth position. Svensk Tandlak T 1946;39(suppl): 1-176. 4 Myllarniemi S. Malocclusion in Finnish rural children: an Figure 5 Likelihood of the pedigree versus penetrance (horizontal axis). Log ar thm'IC epidemiological study of different stages of dental develop- scale of total likelihoods plotted on the vertical axis. Maximum estimate and i ts ment. Doctoral thesis, University of Helsinki, 1970. standard error by quadratic interpolation. 5 Jorgenson RJ. Mandibular prognathism. In: Buyse ML, ed. Birth defects encyclopedia. Cambridge: Blackwell Scien- tific Publications, 1990:1098. 6 Litton SF, Ackerman LN, Isaacson RJ, Shapiro BL. A of class III malocclusion. Am J Orthod pedigree analysis. In the case of an aui tosomal genetic1 970;58:565-77.study dominant gene, there are several ways tio coun- 7 Iwagaki H. Hereditary influence of malocclusion. Am J bi 2627 Orthod Oral Surg 1938;24:328-36. teract or correct for ascertainment I bias 8 Kraus BS, Wise WJ, Frie RA. Heredity and the craniofacial However, if penetrance of a dominant g,ene can complex. Am ] Orthod 1959;45:172-217. 9 Schulze C, Wiese W. Zur Vererbung der Progenie. Fortschr be estimated from a single large pcedigree, Kieferorthop 1965;26:212-29. ascertainment bias plays a lesser role iff subse- 10 Stiles KA, Luke JE. The inheritance of malocclusion due to mandibular prognathism. 7 Hered 1953;44:241-5. quent collection and documentation c)f pedi- 11 Suzuki S. Studies on the so-called reverse occlusion. J gree members is independent of the pheno- Nihon Univ Sch Dent 1961;5:51-8. typic target condition. This can be safely 12 DownsDent ResWG.1928;8:367-79.Studies in the causes of dental anomalies. ] assumed to be the case regarding man dibular 13 Keeler CE. Heredity in dentistry. Dent Cosmos 1935;77:1147. prognathism in European noble lineag 14 Rubbrecht 0. Study of the heredity of the anomalies of the We paid particular attention to noit intro- jaws. Am Orthodont 1939;25:751-79. 15 Stewart RE, Prescott GH. Oralfacialgenetics. St Louis: CV ducing additional bias by cutting the p)edigree Mosby, 1976:96. into manageable parts. The advantag,es of a 16 Penrose LS. The genetical background of common diseases. Acta Genet 1953;4:257-65. single large pedigree are clearly comprromised 17 Rubbrecht 0. L'origin du type familial de la maison de by this procedure. Apart from the prolblem of Habsburg. Bruxelles: van Oost et Cie, 1910. -ed, that 18 Strohmayer W. Die Vererbung des Habsburger Familien- http://jmg.bmj.com/ bias there is a loss of power to be expect typus. Nova Acta Leopoldina 1937;5:219-96. is, variance increase of the MLE. Some experi- 19 Grabb WC, Hodge GB, Dingman RO, O'Neal RM. The mentation with different ways of dividling the 20 ThompsonHabsburg EM,jaw. PlastWinterReconstrRM. SurgAnother1968;42:442-5.family with the pedigree did show, however, that bothi effects Habsburg jaw. J Med Genet 1988;25:838-42. ealth of 21 McKusick VA. Mendelian inheritance in man. 9th ed. Balti- apparently are not substantial: the w( Smore: Johns Hopkins University Press, 1990, No 176700. data provided by this pedigree overrides meth- 22 Sander H. Die Vererbung der Progenie im Hochadel Europas. Inaugural Thesis, Faculty of Medicine, University of odological shortcomings and compromiise.*se. Freiburg, 1992. In summary, the mandibular progrnathism 23 Elston RC, Stewart J. A general model for the genetic on September 28, 2021 by guest. Protected copyright. analysis of pedigree data. Hum Hered 1971;21:523-42. phenotype segregating in related famiilies of 24 Ott J. A computer program for linkage analysis of general European high nobility is determine:d by a human pedigrees. Am] Hum Genet 1976;28:528-9. single autosomal dominant gene of hig~h pene- 25 EdwardsPress, 1972.AWF.ReprintLikelihood.1987.London: Cambridge University trance (0 95). Nevertheless, there is a cc nsider- 26 Greenberg DA. Inferring mode of inheritance by compari- son of lod scores. Am 7 Hum Genet 1989;34:480-6. able variation in expression ranging frc:m pro- 27 Ott J. Analysis of human genetic linkage. Revised edition. truding lower lip to orofacial malfo: rmation Baltimore: Johns Hopkins University Press, 1991.