The Secret of Secrets: Carbonic Anhydrase Concentration in Lizards' Femoral Gland Secretions Is Tuned to Environmental Conditions

Abstract

Among the modalities of animal communication, the chemical channel is the only one that allows signalers and receivers to communicate without being necessarily in the same place at the same time. This asynchrony may influence signal design, as its effectiveness depends on adapting to predictable environmental conditions. Many lizard species scent-mark territories by depositing waxy secretions made of protein-lipid mixtures from specialized epidermal glands. These secretions, once left on a substrate, face environmental fluctuations, and their lifespan depends on the mixture's ability to tolerate such changes. Since some proteins in these secretions are enzymes, we hypothesize they may serve homeostatic functions, enabling the mixture to actively respond to external conditions. Accordingly, we can expect that (1) enzymes remain active in secretions; and (2) their abundance varies across an environmental gradient, tuning the homeostatic ability to the predictable external conditions. We tested these predictions using carbonic anhydrase (CA), an enzyme found in many lizard secretions that regulates pH by catalyzing the reaction between water and carbon dioxide. Protonography confirmed CA activity in proteins extracted from femoral gland secretions from eight Podarcis lizard species and Psammodromus algirus. We also analyzed CA abundance in males (N = 70) from 12 Ps. algirus populations, and its correlation with the environmental gradient (geography, topography, climate). Model comparison revealed a credible relationship between CA concentration and geography as well as bioclimatic variables in Ps. algirus. Taken together, these findings suggest CA plays an enzymatic function that may help stabilize the internal chemical environment of secretions, potentially enhancing the scent-mark's lifespan and effectiveness under varying environmental conditions.

Keywords: chemical communication; environmental factors; enzyme; homeostasis; scent marking.

FIGURE 1

Distribution of the sampling localities of the lizard species used in the study. Left panel: Dots mark the sampling site for the eight Podarcis species; green and orange polygons define the two geographic regions (central and southern Spain, respectively) from which Psammodromus algirus populations were sampled; dark‐gray shaded area represents the distribution of P. algirus (Roll et al. 2017). Right panel: Violin plot of climatic variability across the two geographic regions of P. algirus populations; T _avg = average annual temperature; T _range = temperature mean diurnal range; Prec_annual = total annual precipitation; Prec_CV = Precipitation seasonality (coefficient of variation); all variables were standardized before plotting to allow comparing different scales.

FIGURE 2

Protonography of Podarcis samples compared to SDS‐PAGE and western blot. (A) Protonography (10 s incubation time). (B) SDS‐PAGE stained with Coomassie blue after protonography. (C) Protonography (20 s incubation time). (D) Western blot of the same samples (modified from figure 5 in Mangiacotti et al. (2023)). Lanes numbering is reported above and below gel images. Samples were always loaded in the same order: 1 = molecular weights (kDa); 2 = P. gaigae; 3 = P. erhardii; 4 =  P. bocagei ; 5 = P. melisellensis; 6 = P. liolepis; 7 =  P. carbonelli ; 8 = P. milensis; 9 =  P. muralis . Black arrows mark the position of the CA band in SDS‐PAGE lanes. Gel A, B, and C were cropped and color‐balanced in Adobe Photoshop CS3. Original images are available as (Figure S3).

FIGURE 3

Protonography of Psammodromus algirus samples compared to SDS‐PAGE. (A) Protonography (20 s incubation time). (B) SDS‐PAGE stained with Coomassie blue after protonography. Samples were loaded on the same lane in each gel, from lane 2 to lane 5; lane 1 loaded molecular weights. Black arrows mark the position of the CA band in SDS‐PAGE lanes. Gel A and B were cropped to the zone of interest and brightness enhanced to improve visualization in Adobe Photoshop CS3. Original images are available as (Figure S4).

FIGURE 4

Results from the linear mixed model of the gradient analysis. (A) Model comparison according to the expected long pointwise predictive density (elpd): Points represent elpd difference with the best model (Δelpd), vertical segments represent the one‐standard‐error interval around the mean difference. (B) posterior prediction of the effect of geographic region on CA concentration from the geographic model. (C) posterior prediction of the effect of total annual precipitation (Prec_annual) on CA concentration as predicted by bioclimatic2 model. (D) posterior prediction of the effect of precipitation seasonality (Prec_CV) on CA concentration as predicted by bioclimatic2 model.