Trabecular architecture of the distal femur in extant hominids

Abstract

Extant great apes are characterized by a wide range of locomotor, postural and manipulative behaviours that each require the limbs to be used in different ways. In addition to external bone morphology, comparative investigation of trabecular bone, which (re-)models to reflect loads incurred during life, can provide novel insights into bone functional adaptation. Here, we use canonical holistic morphometric analysis (cHMA) to analyse the trabecular morphology in the distal femoral epiphysis of Homo sapiens (n = 26), Gorilla gorilla (n = 14), Pan troglodytes (n = 15) and Pongo sp. (n = 9). We test two predictions: (1) that differing locomotor behaviours will be reflected in differing trabecular architecture of the distal femur across Homo, Pan, Gorilla and Pongo; (2) that trabecular architecture will significantly differ between male and female Gorilla due to their different levels of arboreality but not between male and female Pan or Homo based on previous studies of locomotor behaviours. Results indicate that trabecular architecture differs among extant great apes based on their locomotor repertoires. The relative bone volume and degree of anisotropy patterns found reflect habitual use of extended knee postures during bipedalism in Homo, and habitual use of flexed knee posture during terrestrial and arboreal locomotion in Pan and Gorilla. Trabecular architecture in Pongo is consistent with a highly mobile knee joint that may vary in posture from extension to full flexion. Within Gorilla, trabecular architecture suggests a different loading of knee in extension/flexion between females and males, but no sex differences were found in Pan or Homo, supporting our predictions. Inter- and intra-specific variation in trabecular architecture of distal femur provides a comparative context to interpret knee postures and, in turn, locomotor behaviours in fossil hominins.

Keywords: Gorilla; Pan; Pongo; bipedalism; functional morphology; human; knee; locomotor behaviour.

FIGURE 1

(a) Ligament attachments of the knee showing inferior view. Gold, anterior cruciate ligament; blue, posterior cruciate ligament; green, lateral collateral ligament; red, medial collateral ligament; brown, lateral meniscus; purple, medial meniscus. The medial collateral ligament inserts at the medial epicondyle of the femur and attaches along the medial border of the tibial plateau and on the medial surface of the tibial shaft. The lateral collateral ligament inserts on the lateral femoral epicondyle and attaches on the head of fibula. (b) Anterior cruciate ligament position during forward tibia and backward femur movements. (c) Posterior cruciate ligament position during backward tibia and forward femur movements. Cruciate ligaments position during (d) neutral knee position; (e) medial knee rotation; (f) lateral knee rotation. The anterior cruciate ligament arises from the anterior intercondylar space on the tibial plateau, runs upwards and posteriorly and attaches on the inside of the lateral condyle of the femur. The posterior cruciate ligament arises from well back on the posterior intercondylar space, runs upwards and anteriorly and attaches on the inside of the medial condyle. (g) Surface models of right distal femur of Homo, Gorilla, Pan and Pongo. Showing more elliptical shape of lateral condyle, square outline of distal surface and high lateral lip in humans compared to great apes.

FIGURE 2

Processing steps of distal femur of a Homo specimen showing in inferior view. (a) Original high‐res microCT image. (b) MicroCT image segmented by MIA (Dunmore et al., 2018). (c) MaskSeg defined by Medtool (Pahr & Zysset, 2009) showing the distinction between inner trabecular area and outer cortex. (d) Outer canonical atlas representing sample's mean size, position and external right distal femur morphology computed by cHMA (Bachmann et al., 2022). (e) Inner mesh representing rBV/TV distribution in human sample computed by cHMA (Bachmann et al., 2022).

FIGURE 3

Species mean models of rBV/TV distribution in the patellar articular surface and femoral condyles of the distal femur of Homo, Gorilla, Pan and Pongo. Vertical lines through the inferior view mean models show where the slices are positioned. Cross sections were positioned in the middle of lateral and medial condyle. (a) Approximate patellar and condyle position during flexed (90°) and extended knee position in humans in medial view; (b) approximate condyle position during flexed and extended knee position in great apes. L, lateral; M, medial.

FIGURE 4

Species mean models of rBV/TV distribution under the presumed insertion of cruciate ligaments of the distal femur of Homo, Gorilla, Pan and Pongo. Vertical lines through the medial and lateral mean models show where the slices are positioned. Cross sections were positioned in the middle of presumed insertions of cruciate ligaments. (a) Anterior cruciate ligaments during neutral knee position in the anterior view.

FIGURE 5

Species mean models of rBV/TV distribution under the presumed insertions of collateral ligaments of the distal femur of Homo, Gorilla, Pan and Pongo. Vertical lines through the medial and lateral mean models show where the slices are positioned. Cross sections were positioned in the middle of presumed insertions of collateral ligaments. (a) Lateral collateral ligaments attachments. L, lateral; M, medial.

FIGURE 6

Species mean models of rBV/TV distribution under the presumed gastrocnemius muscle attachments of the distal femur of Homo, Gorilla, Pan and Pongo. Vertical lines through the posterior mean models show where the slices are positioned. Cross sections were positioned in the middle of presumed gastrocnemius attachments. (a) Gastrocnemius attachments. L, lateral; M, medial.

FIGURE 7

Species mean models of DA distribution in the patellar articular surface (marked with square) and in the presumed insertions of gastrocnemius muscle attachments (marked with circle) of the distal femur of Homo, Gorilla, Pan and Pongo. Vertical lines through the inferior view mean models show where the slices are positioned. Cross sections were positioned in the middle of lateral and medial condyle. L, lateral; M, medial.

FIGURE 8

HMA and segmented models showing high DA concentration in the lateral/medial condyle and under the presumed insertions of gastrocnemius muscle attachments (marked with circle) of the distal femur of (a) Homo and (b) Gorilla individual. Vertical lines/squares show where the slices are positioned. L, lateral; M, medial.

FIGURE 9

PCA of rBV/TV distribution in the distal femur of Homo, Gorilla, Pan and Pongo showing separation among studied taxa. Thresholding models represent high 30% of the range of rBV/TV values for negative and positive PCAs. Models demonstrate the highest loading causing the separation between humans (negative PC1 − 3SD) and great apes (positive PC1 + 3SD); between Pongo, Pan (negative PC2 − 3SD) and Gorilla (positive PC2 + 3SD). L, lateral; M, medial.

FIGURE 10

PCA of DA distribution in the distal femur of Homo, Gorilla, Pan and Pongo showing separation among studied taxa. Thresholding models represent high 30% of the range of DA values for negative and positive PCAs. Models demonstrate the highest loading causing the separation between humans (positive PC1 + 3SD) and great apes (negative PC1 − 3SD); between Gorilla, Pan (positive PC2 + 3SD) and Pongo (negative PC2 − 3SD). L, lateral; M, medial.

FIGURE 11

PCA of rBV/TV and DA distribution in the distal femur of Gorilla and Pan. (a) PCA of rBV/TV distribution of Gorilla and Pan showing separation between sexes in Gorilla and no separation in Pan. (b) PCA of DA distribution of Gorilla and Pan showing separation between sexes in Gorilla and no separation in Pan. Thresholding models represent high 30% of the range of rBV/TV and DA values for negative and positive PCs. Models demonstrate the highest loading causing the separation between female and male gorillas. L, lateral; M, medial; F, female; M, male.