vendredi 27 février 2015

Cnidaria: biological control over calcification

Environmental and Evolutionary implications


Up to recently (and still accepted in many places) calcification in the Cnidaria was admittedly a poorly controlled process contrasting with the long standing agreement regarding the biochemically driven crystallization responsible for calcification of mollusks shells. In the latter the organisms exert a permanent full control over mineralogy, sizes, shapes and spatial arrangements of the shell building units whereas in corals crystallization of fibers was considered as simply biologically induced (Lowenstam 1981, Veis 2005). Acceptance of such a weak biological control led to methodological and practical consequences regarding two major research domains: 
1 - no attempts were made to explain the origin of the “vital effect” that restricts the use of Cnidaria skeletons as environmental archives to the rare “calibrated” species; 
2- the potential of coral fibrous structures as evolutionary tracers was largely denied. In the reference Treatises dealing with fossil corals priority was given to morphological characters, resulting in the discrepancy between the classical evolutionary trees (e.g. Wells 1956 and followers) and the growing number of data produced by two decades of molecular methods applied to coral polyps.

This three-fold article summarizes:

  • The sequence of physical characterization methods progressively revealing that coral skeletons are also produced through a biologically controlled process;
  • The consequences of this biological control regarding the use of coral skeletons as environmental archives when chemical or isotopic measurements are made without taking into account the microstructural patterns;
  • The taxonomic and evolutionary potential of the microstructural investigations when carried out at the relevant levels.

1- Evidence for a full biological control over skeleton crystallization in the Cnidaria: similarities between compositions and distributions of organic contents within calcareous skeletons and their resulting ultra-structures


Evidence for a biological control over Ca-carbonate crystallization was first recognized in the mollusk shells when, in middle of the 19° century, optical microscopes were used not only to describe the tiny morphological patterns of the shells but when Bowerbank (1844), Carpenter (1845) and then numerous others began to pay attention to the structure of the mineral phase forming the shells. Under the microscope mollusk shells were shown to be built not by a simple and unorganized aggregation of calcium carbonate grains but by layers of well defined calcareous units. To summarize, shells of a given species are regularly made of at least two layers of morphologically and sometimes mineralogically distinct skeletal units. 


1a - In the skeletons of scleractinian corals, by far the most popular and geologically important calcifying Cnidaria (they are the major reef builders), such a dual organization remained controversial up to the last decades of the 20th century. As early as 1896 M. Ogilvie clearly established the taxonomic value of the spatial arrangements of the fiber fascicles and also coined the term “centres of calcification”. But these centers were defined as “the points from which fibres diverge”, casting doubts about their actual nature. It is only a century later that, taking advantage of scanning microscopy and electron microprobe chemical measurements, it was shown that centers of calcification and fibers were clearly distinct structures differing with respect to both morphology (Fig. 1) and chemical compositions (Fig. 2) of the crystalline units [C136-C182]. Conclusively, skeletons of the Scleractinia are also built by two distinct and superposed mineralization areas.

 Figure 1 - Calicinal view (a) and closer view (b) of a scleractinian coral septum. The crest line (d) is built by distinct units (the so called “centers of calcification”) formed of agglutinated micrometer sized grains whereas fibers are built by superposed layers of calcareous rods (e) of about 2-3 µm long. Scheme (f) and microscopic thin section (c-g) show that the whole skeleton is built through a regular stepping growth mode. 


Presence of an organic component associated with the calcareous skeletons also contributes to make the scleractinian skeletogenesis essentially comparable to the biochemically driven process running during mollusk shell formation: both organic components exhibit very high concentrations in acidic amino acids (Mitterer 1978, Weiner 1979). Quantitative evaluation using thermogravimetric analysis (TGA) leads to comparable results : in coral skeletons content of the non-mineral phase is in the 2-3% in weight range with about half of this weight represented by water [C169]. 


Figure 2 - Example of TGA measurement (a), UV fluorescence (b) and synchrotron based XANES mapping providing evidence for the presence and distribution of organic components within coral skeletons. 2a: The red line show the weight loss during heating. Note the rapid loss around 300°C. IR absorption has shown that this rapid loss was due to water emission (dotted blue line). In parallel, IR absorption of CO2 also reveals the sequence of thermal decay of the organic components (colored sectors); 2b: UV epi-fluorescence of a coral septum showing the higher organic content in the median area of the septum (zone of initial mineralization); 2c: XANES mapping of organic sulfur (sulfated proteins and polysaccharides) showing the higher organic content in the median area of the septum and the layered distribution of organics in the fibers.


But it is only when synchrotron based XANES mapping has made clear a layered distribution of these organic components within the coral skeletons that their stepping growth process was obvious [C151]. Thus in both corals and mollusks the mineralizing epithelium (mollusk mantle and coral basal epithelium) are using the same method to control crystallization of their skeletal units, whatever their sizes, shapes and spatial arrangements.

1b - The common granular infra-micrometric pattern of mineral phases in biocrystals.
What provides a conclusive evidence of similarity between molluscs and scleractinia regarding the the biomineralization process is the remarkable resemblance of their mineral phases observed at the inframicrometric level. Using Atomic Force Microscopy (AFM)  the mineral component forming the aragonite fibers of the Scleractinia (whose polarized light emphasizes the single-crystal behavior on microscopic thin sections; Fig. 3a) exhibits a reticulated structure resulting from a close contact between irregular round shaped grains (Fig. 3b-c). Another well known Cnidaria, the Corallium rubrum, a member of the Octororalla group, built calcitic ramified axes (Fig. 3d-e) whose polished and etched sections also reveal the layered growth mode (Fig. 3f), a pattern that is also visible in the intracellular spicules that form a protective cortical layer at the periphery of the ewes (Fig. 3g-h). In both axes and spicules the mineral component exhibit this reticular structure that makes obvious the specific mode of crystallization of calcium carbonate in the biocrystals.

That this mode of crystallization is widely shared is exemplified by comparing the infra-micrometric patterns in the aragonite fibers of the Scleractinia (Fig. 3b-c), the layered calcite growth units of Corallium rubrum (Fig. 3i) [YD242] and the grains from the calcite prisms of the shell of Pinctada margaritifera in which the organic cortex covering the grains is evidenced by phase contrast imaging at the AFM (Fig 3j-k).


Figure 3 - The common layered growth mode and granular infra-micrometric pattern in the calcareous biocrystals from a Scleractinia (a-c), an Octororalla (d-i, Corallium), and the Pinctada prisms (j-k, Mollusca Pelecypod). 3a: Fibrous fan-system forming the major part of the Scleractinia skeletons thin section polarized light. Note the synchronic growth layers visible at the upper part of the view. 3b-c: Fine structure of the aragonite crystals in the Scleractinia fibers: AFM height image (b) and phase contrast (c). 3d-e: Skeletal axis of a Corallium rubrum colony with expanded polyps (b). 3f-h: Both axes (3f) and intracellular spicules (3g-h) exhibit a layered growth mode, with a typical reticulated structure of the magnesian calcite (3i). 3j-k: Granular structure of the calcite in the prisms of Pinctada margaritifera. The phase-contrast mode emphasizes the organic-rich coating of the grains.


An about ten years extensive investigation [C244] has revealed that a layered growth and crystallization model [C251] may account for the micrometric and infra-micrometric calcification patterns observable not only among all invetebrate groups but also for the calcareous structures developed by Vertebrates (e.g. otoliths and egg shells).

2. The use of coral calcareous structures as environmental archives with focus on the consequences of their biochemically driven mode of crystallization

Up to the middle of the 20th century the calcareous structures produced by many organisms were used as indicators for the characterization of ancient environments through a purely qualitative approach. Corals, for instance, were commonly considered as indicative of shallow waters with inter-tropical temperatures. A major methodological breakthrough was introduced by H.C. Urey in the 1950’s taking advantage of his in-depth understanding of isotopic fractionation (Nobel prized 1934 for discovery of Deuterium).

2a: The concept of “vital effect”
Urey first produced (1947) a mathematical expression for the difference that would be established during chemical reactions between proportions of the isotopes of a given chemical element before it entered into the reaction and the isotope ratio of this same element after the reaction, i.e. within the new molecules resulting from the reaction. Urey showed that the new isotope ratio for this element varies as a function of the temperature at which the reaction has occurred. He suggested that, reciprocally, the measure of isotopic ratios in a given material could provide an indication of the temperature at which this material was formed. Urey stated: “I suddenly found myself with a geological thermometer in my hands.”
By carrying out a series of in vitro Ca-carbonate precipitations McCrea (1950) experimentally checked Urey’s calculations and, in 1951 the result of the first practical application of the isotope-based paleo-temperature reconstruction was published (Geol. Soc. America Bull. 62: 399-416), rapidly extended to numerous investigations and papers that triggered the rise of the “isotopic paleontology” (Wefer & Berger 1991).

But these measurements resulted in a strange conclusion. If isotopic ratio and also chemical partitioning of trace elements in the skeletons undoubtedly vary according to temperature change, it is remarkable that each species exhibits its own slope when plotting temperatures versus the corresponding chemical concentrations or isotopic fractionations measured along the skeleton growth direction. An extensive study of skeletons produced by several coral species from the Pacific Ocean allowed Weber & Woodhead (1972) to illustrate this paradoxical result using natural samples and numerous experimental cultivation fully confirmed this taxonomy-linked phenomenon.

Figure 4 - A series of punctual measurements of isotopic ratio carried out on skeleton of a Porites colony (a-b) compared to a reference (here the powder from the Pee Dee Belemnite) allow a normalized profile of variation (c). But due to the “vital effect” this profile can be transformed to a temperature profile only if the used species has been previously “calibrated”, i.e.  its own slope of sensitivity to temperature variation carefully established (d : redrawn from the original figure of the distinct slopes obtained from a series of different coral species by Weber & Woodhead).

Such a result fully justified the somewhat intriguing remark formulated by Urey in his seminal 1951 paper: “we may ask whether there is a vital effect” (p. 451), a remarkable premonition at that time as no biochemical mechanism was available to support this idea. Since this period a “calibration” process is a pre-requisite for using the calcareous structure built during its life by any organism as environmental archive. Even more remarkably, detailed investigations focused on the  mode of growth of calcareous skeletons carried out during the last decades have shown that the Urey’s concept of “vital effect“ was valid not only at the species level but much more precisely, for each of the distinct biomineralization areas whose superposition contributes to skeleton construction.

2b: The simultaneous record of distinct isotopic or chemical signals in calcareous skeletons: evidence of a link between “vital effect” and biocrystallization

That coral skeletons are built by two distinct biomineralization areas was first confirmed by specific preparation (polished and etched surfaces observed at the Scanning Electron Microscope) leading to formal validation of the old concept of “urseptum” (Volz 1896; see also 2c). Emergence of physical instruments allowing isotopic or chemical measurements to be carried out with spatial resolution in the 10 micrometer range brought interesting information about the properties of these microstructurally distinct areas.

The Lophelia case
This species was the first to be studied by mean of high resolution instruments, taking advantage of a very simple spatial arrangement of the two distinct mineralizing areas. The upper area at the top of the septa (Fig. 5a-c) is made of densely packed spheroidal units (Fig. 5d). As this sommital growth line is rapidly covered during the growth process by two thick layers of fibrous tissue, typically symmetrical on both sides. Here we observe a typical example of primary “urseptum” followed by production of reinforcing fibrous structure (Fig. 5e-f).

Figure 5 - Branching pattern (a), single calice (b) oblique view of a septum (c) and microstructure of the upper growing area (d) in Lophelia pertusa; microscopic thin section (e) showing the upper growth area between the layers of fibres and microstructural pattern of the (f ) area (secondary electron microscopy); g: measurements carried out with  high resolution sampling instrument (CAMECA 1270 Secondary Ion Mass Spectrometer) allow clear recognition of the distinct isotope fractionation in these two micro-structurally distinct areas (redrawn from C. Rollion-Bard, 2011).

2c: A practical consequence of the simultaneous recording of distinct signals in the two distinct mineralizing areas of coral skeletons: measurements irrespective of the microstructural patterns result in highly scattered numerical values and weakened significance of the environmental signal.

It is of common practice (specifically when using automatic machines) that sampling is made linearly on cores drilled from a coral colony Fig. 6a, b). This results in a very high dispersion of fractionation values (Fig. 6c: example from a study by C. Rollion 2001). Worth to note that dispersion of the measured values in this example from a Porites colony is in the same range (several delta units) than previously observed in the Lophelia (Fig. 5g). Remark must be made that the Lophelia values are forming two dense clusters clearly distinct because measurement were carried out by taking into account the microstructure of the skeleton. Just the reverse in the Porites case. Fine scale observation of the Porites microstructure allows to understand that in a “blind sampling” the spot can be entirely located on fibres (Fig. 6g1) or almost entirely on “centres of calcification” (Fig. 6g4) with in between any proportion of the two distinct mineralization areas (Fig. 6g2-g3).
Undoubtedly, measurements carried out following the usual sampling process provide investigators with basically mixed values, leading to weakening of the true environnemental signals.  

Figure  6 -  “Blind sampling” on Porites skeleton  a-b: Longitudinal section in a Porites skeleton. SEM view (a) and microscopic thin sections (b: polarized light): c: dispersion of measurements made without taking into account the microstructural patterns. Note that dispersion amplitude of several delta units; d, e, f, g: Detailed observation of the Porites skeleton allows understanding of the dispersion. Location of the sampling spot can includes various proportion of the two distinct areas, the deltas of which are highly distinct, as previously shown.

Rollion-Bard C. (2001) Variability of oxygen isotopes in Porites corals: development and implications of stable isotopes (B, C and O) microanalysis by ion microprobe. Thèse, Institut national polytechnique de Lorraine, Nancy, 165 pp.

Rollion-Bard C. (2011) Processus de biominéralisation dans les carbonates biogéniques et impact sur les traceurs environnementaux. HDR, 2 Décembre, CRPG Nancy.