Piezometric architecture and the geometry of plasticity: a proposed electromechanical model of the Fibrosis–Plasticity axis

1 Introduction

Tissues in terrestrial organisms bear continuous gravitational load, and in bipedal humans this load is distributed asymmetrically and posture-dependently through the body’s mechanical architecture (Rowell, 1974; Smith et al., 1994). Clinical physiology characteristically reduces pressure to scalar hemodynamic quantities measured at a reference height, which serves its purpose but does not describe the spatial organisation of mechanical load in tissue (Guyton and Hall, 2016).

Fibrosis is conventionally described as a wound-healing response that fails to terminate, with myofibroblast persistence, altered matrix turnover and progressive stiffening as its principal features (Rockey et al., 2015; Wynn and Ramalingam, 2012; Henderson et al., 2020). That description is well supported. It does not by itself specify why, absent discrete injury, some tissues progress towards fixation over years while others retain remodelling capacity, or whether the spatial organisation of chronic load contributes to that divergence.

1.1 Mechanical terminology used in this manuscript

Because the argument moves between scales, the mechanical quantities are distinguished explicitly and are not used interchangeably.

• Load—the externally applied force system acting on the organism or a region of it, including gravitational load.
• Load geometry—the spatial arrangement by which load is distributed through the body’s mechanical architecture. This is the organism-level quantity the framework concerns.
• Stress—force per unit area within tissue (Pa).
• Strain—the resulting local dimensional change, dimensionless.
• Hydrostatic pressure—the scalar isotropic component of the stress tensor, and the quantity measured clinically as blood pressure. Hydrostatic pressure alone produces no shear and is not the quantity that generates the electromechanical phenomena discussed here.
• Deformation—the geometric change in matrix configuration arising from the deviatoric (shear) components of stress, and the immediate cause of the electrical phenomena under discussion.

The term piezometric is used in this manuscript descriptively, for the spatial distribution of mechanical load. It is not used as a claim about hydrostatic pressure, and the proposed chain does not run from hydrostatic pressure to electrical signal. It runs from load geometry to local stress and strain to matrix deformation to electrical phenomena.

1.2 Terminology introduced in this work

The following constructs are introduced or defined here and are not established entities in the existing literature: the Fibrosis–Plasticity Axis (FPA) as a named conceptual state-space; the Geometric Origin of the Body (GOB); the Geometric Origin Displacement Index (GODI); and the Geometric Optimal Pressure Load (GOPeL). Where earlier work by this author used the variant GOBen, that term is superseded by GOB.

The following are drawn from the existing literature and cited accordingly: tensegrity as a model of cellular and tissue mechanical organisation (Ingber, 2008); fascial continuity and force transmission (Schleip et al., 2012; Langevin, 2021); strain-generated potentials in bone and connective tissue (Athenstaedt, 1974; Fukada and Yasuda, 1957; Minary-Jolandan and Yu, 2009; Abdollahi et al., 2019; Giraud-Guille, 1988; Anderson and Eriksson, 1968; Pienkowski and Pollack, 1983; Otter et al., 1985; Ahn and Grodzinsky, 2009; Halperin et al., 2004; Kwon and Cho, 2022; Vasquez-Sancho et al., 2018; Abdollahi et al., 2018); and mechanotransductive control of matrix production (Humphrey et al., 2014; Hinz, 2015; Zhao et al., 2022).

1.3 The electromechanical proposal, stated as nested hypotheses

Piezoelectricity in dry bone was reported in 1957 (Fukada and Yasuda, 1957), and the electrical behaviour of wet collagen was examined shortly afterwards precisely because tendon and bone are bathed in high-conductivity ionic fluid in vivo (Anderson and Eriksson, 1968). Whether a solid-state response persists and is biologically consequential under full physiological hydration has been disputed since, with electrokinetic streaming potentials advanced as the dominant source of macroscopically measured potentials in fluid-saturated tissue (Pienkowski and Pollack, 1983; Otter et al., 1985; Ahn and Grodzinsky, 2009) and flexoelectricity more recently proposed as a further contributor (Vasquez-Sancho et al., 2018). A framework that simply asserted collagen piezoelectricity as its physical foundation would inherit this unresolved dispute as an unstated assumption.

The proposal is therefore stated as three successively nested hypotheses. Each is separately testable, and each may fail without falsifying the one above it.

1.3.1 H1 — Broad electromechanical hypothesis

Deformation of collagen-rich extracellular matrix generates local electrical phenomena—comprising electrokinetic (streaming) potentials, piezoelectric contributions, and potentially flexoelectric contributions—which may modulate the behaviour of matrix-producing cells and thereby contribute to remodelling outcomes.

1.3.2 H2 — Solid-state hypothesis, nested within H1

A solid-state component of these phenomena contributes at physiological hydration. Not requiring fluid flow for its generation, such a component would be expected to persist when electrokinetic contributions are reduced by limiting interstitial fluid movement—though, as §8.3 sets out, no available manipulation reduces fluid movement without also altering hydration and the distribution of deformation.

1.3.3 H3 — Piezoelectric specificity hypothesis, nested within H2

The solid-state component is specifically piezoelectric—polarisation coupling to strain in a non-centrosymmetric molecular arrangement—rather than flexoelectric, in which polarisation couples to the strain gradient and non-centrosymmetry is not required.

The distinction between H2 and H3 is not a refinement but a necessary separation. Flexoelectricity is a solid-state phenomenon, so an experiment showing only that a potential persists when fluid movement is reduced bears on H2 and leaves H3 undecided. Section 8 specifies separate criteria for each.

H1 is the hypothesis on which the framework’s biological relevance depends. H2 and H3 determine what kind of physical mechanism is involved, and bear on whether the phenomenon would persist in dehydrated or fluid-depleted matrix—a question with direct relevance to fibrotic tissue, in which water content is altered.

1.4 Aim

The aim is to state, as a chain of separable and gradeable propositions, how chronic load geometry might bias tissue towards fixation (Figure 1); to identify which links are novel; and to specify how they could be refuted.

2 Terminology and epistemic status

Table 1 gives definitions together with an explicit epistemic status. No entry should be read as a validated measure.

TABLE 1. Terminology introduced or used in the present framework, with epistemic status.

2.1 What constitutes the fibrosis–plasticity axis

The FPA is a proposed conceptual continuum, not a measured variable and not an established biological axis. It asks one question of otherwise separately studied observables: how much of a tissue’s remodelling capacity remains recoverable?

The following are offered as candidate dimensions proposed to represent the FPA. They constitute the author’s proposed operationalisation. No factor-analytic or biological validation has been performed to show that they describe a single latent axis, and the proposal that they do is itself a hypothesis.

• Matrix turnover capacity—rates of collagen synthesis and degradation; MMP/TIMP balance.
• Crosslink maturity—proportion of crosslinks that are no longer reducible.
• Myofibroblast state—presence, persistence and reversibility of the activated contractile phenotype (Rockey et al., 2015; Hinz, 2015; Tomasek et al., 2002).
• Matrix architecture—fibre alignment, density and spatial isotropy.
• Demonstrated reversibility—observed capacity to return towards baseline after a perturbation is withdrawn.

Bulk tissue stiffness, elastography values, tissue impedance and clinical range-of-motion measures are treated as candidate correlates, not as dimensions of the axis.

No validated composite FPA score exists. Constructing one would require agreement on which dimensions are measured, at what scale, with what weighting, with demonstrated test–retest reliability, and with evidence that the dimensions covary as a single construct. Until then the FPA functions as a framing device and a hypothesis generator.

2.2 Candidate operationalisation of GODI and GOPeL

No formula is proposed. What can be indicated are the measurement domains a future operationalisation would need to sample, given here as illustrative candidate measures, not as components of a proposed score: static load distribution; segmental alignment; time to return to baseline configuration after a standardised perturbation; sustained compensatory muscle activation at rest; tolerance of standardised orthostatic challenge; and regional tissue compliance.

For GOPeL, corresponding candidate criteria would be absence of sustained rather than task-appropriate compensatory activation, full return of segmental configuration after perturbation, and absence of accumulated residual strain across a loading cycle.

Variable selection, weighting, units, inter-rater and test–retest reproducibility, and criterion validation against any independent marker of tissue plasticity remain to be done. Any statement of the form “higher GODI is associated with reduced plasticity” is accordingly a hypothesis, not a reported relationship.

3 Load geometry as a spatial field

3.1 Beyond scalar blood pressure

Clinical physiology operationalises arterial pressure as a value measured at heart level, for reasons of standardisation. The circulation is nonetheless a closed, gravity-exposed system in which hydrostatic pressure is positional: pressure at the ankle and at the head differ substantially and predictably with posture (Convertino, 2002). A single measurement is a projection of a three-dimensional field.

This observation motivates the framework’s attention to spatial distribution, but the causal chain proposed below does not run through hydrostatic pressure. As set out in §1.1, the quantity that generates electromechanical phenomena is matrix deformation arising from the deviatoric components of stress.

3.2 Chronic load asymmetry: The supported claims

Where load distribution is chronically asymmetric, several consequences have empirical support. Interstitial fluid movement, and with it solute delivery and clearance, is sensitive to local pressure gradients and interstitial pressures (Swartz and Fleury, 2007; Levick and Michel, 2010). Vessels and connective tissues remodel in response to sustained mechanical loading (Humphrey, 2002). Cells alter gene expression and matrix production in response to their mechanical environment through well-characterised pathways (Humphrey et al., 2014; Hinz, 2015; Zhao et al., 2022; Tomasek et al., 2002).

In loose connective tissue specifically, applied stretch produces measurable fibroblast cytoskeletal remodelling (Langevin et al., 2005), and moderate stretch has been shown to decrease soluble TGF-β1 and type-1 procollagen in mouse subcutaneous connective tissue in ex vivo and in vivo models (Bouffard et al., 2008). This last finding is worth stating in the direction the original reports it. Mechanical loading of connective tissue modulates fibrogenic signalling, but the direction of that modulation is context-dependent rather than monotonic: moderate stretch was anti-fibrotic in this model, whereas sustained or high-magnitude loading in other systems is pro-fibrotic. The framework proposed here concerns chronic directional bias in load distribution, not loading magnitude as such, and it should not be read as predicting that more strain produces more matrix.

That chronic asymmetry of whole-body load geometry is itself a driver of progression towards fibrotic fixation, absent injury, is not established. It is a proposition of this framework.

4 Geometric Optimal Pressure Load (GOPeL)

GOPeL is proposed as the state in which internal load is distributed such that structures bear gravitational load with minimal sustained compensatory activation and maximal reversibility. As proposed, it would be characterised by load distribution that varies appropriately with posture, activity and respiration without sustained directional bias; recruitment of postural musculature for task demands rather than continuous bracing; and return of tissue configuration to baseline after perturbation without accumulated residual strain.

We further propose that GOPeL is not equivalent to normotension, not reducible to hemodynamic metrics, and unlikely to be achievable by local correction alone. These are conjectures consistent with the framework rather than findings, and GOPeL is not presented as a physiological optimum that has been demonstrated.

5 Deformation-generated electrical phenomena in collagenous tissue

5.1 Mechanisms and their current standing

Three mechanisms are candidates, and their standing differs. Each is described below in terms of what the cited studies measured, and in what preparation.

5.1.1 Piezoelectricity

Fukada and Yasuda reported piezoelectricity in dry bone in 1957 (Fukada and Yasuda, 1957). A non-centrosymmetric molecular arrangement of type I collagen provides the structural basis for a shear piezoelectric response, the reported coefficients belonging to a polar point group (Athenstaedt, 1974; Anderson and Eriksson, 1968). At the nanoscale, Halperin and colleagues applied piezoresponse force microscopy (PFM) to human bone and reported piezoelectric response in both wet and dry specimens, with calculations accounting for field inhomogeneity under the tip and screening in the conductive sample (Halperin et al., 2004). Minary-Jolandan and Yu, using the same technique on isolated collagen fibrils, reported piezoelectric heterogeneity coinciding with the gap and overlap periodicity of the fibril (Minary-Jolandan and Yu, 2009). Kwon and Cho subsequently reported that this gap–overlap heterogeneity is diminished in osteogenesis imperfecta collagen and proposed a role in intrafibrillar mineralisation (Kwon and Cho, 2022).

Four qualifications apply to this group of findings, and they should not be merged.

First, the measurements are at nanometre scale under a scanning probe tip, not at tissue scale under physiological loading. Second, the gap-region heterogeneity and the mineralisation proposal derive from fibril-level work and do not by themselves establish tissue-level electromechanical signalling. Third, the “wet” condition in probe microscopy is not equivalent to fully perfused tissue under physiological ionic and flow conditions.

Fourth, and most consequentially for the hypothesis structure adopted here, PFM does not by itself establish that a measured response is piezoelectric. The technique delivers a voltage through a sharp tip, which necessarily produces a spatially inhomogeneous electric field in the sample; Abdollahi and colleagues showed theoretically and experimentally that this configuration generates converse flexoelectricity, and that large effective piezoelectric coefficients can accordingly be measured in materials that are not piezoelectric at all (Abdollahi et al., 2019). The converse confusion has also been demonstrated: apparent flexoelectric behaviour can arise from piezoelectricity (Abdollahi et al., 2018). The nanoscale evidence in this group therefore supports the presence of a solid-state electromechanical response—H2 — considerably more strongly than it supports a specifically piezoelectric one—H3. This is a principal reason for separating the two hypotheses rather than treating collagen piezoelectricity as the established foundation of the framework.

5.1.2 Electrokinetic streaming potentials

Anderson and Eriksson examined wet collagen specifically because tendon and bone are bathed in high-conductivity ionic fluid in vivo, and reported behaviour differing from dry collagen (Anderson and Eriksson, 1968). Pienkowski and Pollack demonstrated that stress-generated potentials in fluid-saturated human and bovine cortical bone vary with the conductivity, ionic strength and viscosity of the saturating fluid, with consistent polarity reversal at high NaCl concentration—a pattern they interpreted as consistent with streaming-potential theory (Pienkowski and Pollack, 1983). Otter and colleagues reported evidence for different dominant sources in wet and dry bone (Otter et al., 1985). Ahn and Grodzinsky’s critical review concluded that streaming potentials had largely displaced collagen piezoelectricity as the accepted account of strain-generated potentials in hydrated bone (Ahn and Grodzinsky, 2009).

5.1.3 Flexoelectricity

Vasquez-Sancho and colleagues measured flexoelectric coefficients in bone and in hydroxyapatite, and calculated, from those measured coefficients, that strain-gradient-generated polarisation near a crack tip could reach magnitudes theoretically sufficient to induce osteocyte apoptosis and thereby initiate crack healing (Vasquez-Sancho et al., 2018). The distinction matters: the coefficients are measurements, the biological consequence is a calculation, and no cellular response to flexoelectrically generated fields in bone was demonstrated in that work.

The two solid-state mechanisms are distinguished by their constitutive relations, not by loading configuration. Piezoelectric polarisation is proportional to strain and requires a non-centrosymmetric structure; flexoelectric polarisation is proportional to the strain gradient and, because any inhomogeneous deformation locally breaks inversion symmetry, occurs in dielectrics of any symmetry. It follows that in a material which is both non-centrosymmetric and mechanically heterogeneous—as collagenous tissue is—no ordinary loading configuration isolates either mechanism. Bending does not produce a purely flexoelectric response, since the local strain at each point still drives a piezoelectric one; and nominally uniform extension does not produce a purely piezoelectric response, since strain gradients arise at fibre boundaries, mineral–matrix interfaces, pores and specimen ends—and collagenous tissue is hierarchically organised at precisely these scales, with fibril orientation changing systematically between lamellae (Giraud-Guille, 1988). The two contributions can also interfere constructively or destructively depending on geometry. Separating them is therefore a matter of exploiting differences in scaling—flexoelectric contributions grow as characteristic feature size falls—rather than of choosing a loading mode, and it remains an unsolved problem in this tissue (§8.3).

These three are not mutually exclusive; a measured signal may contain contributions from all, and the coupling between piezoelectric and streaming contributions is itself an open question.

The position taken here is that of §1.3. H1 requires only that load-dependent local electrical phenomena of some origin exist and modulate cell behaviour. H2 asserts a solid-state component. H3 asserts that this component is piezoelectric rather than flexoelectric. The umbrella term SGP is used where the origin is not at issue, with the terminological caveat noted in Table 1.

5.2 Systemic mechanical coupling with distributed local transduction

The framework does not propose that electrical signals propagate over long distances through tissue. Extracellular fluid is a conductive ionic medium, and no argument in what follows depends on any particular quantitative account of attenuation, because no long-range electrical propagation is claimed.

The claim about systemic organisation is mechanical:

• Mechanical load is distributed through mechanically continuous tissue over long distances. Collagen-rich matrix deforms at many sites along that distributed load path. Where matrix deforms, local strain-generated potentials arise. What is distributed is therefore deformation; the electrical phenomena are local and multi-focal, arising independently wherever deformation occurs.

Under this formulation, a chronic bias in trunk load geometry may be associated with altered local electromechanical conditions at distant sites without any field propagating from trunk to periphery. The systemic element is mechanical continuity; the electrical element is local.

Several quantities central to this proposal are not established in vivo for non-osseous connective tissue: the magnitude of physiologically occurring SGPs; their spatial range beyond the deforming volume; their temporal persistence under sustained as opposed to cyclic load; and whether they reach magnitudes at which resident cells respond. Table 2 stratifies the available evidence (Figure 2).

TABLE 2. Evidence stratification for the electromechanical component.

5.3 Candidate sites

Bone is the tissue in which load-dependent electrical phenomena and remodelling have been studied most closely (Fukada and Yasuda, 1957; Pienkowski and Pollack, 1983; Frost, 1987; Duncan and Turner, 1995). Skeletal elements are mechanically coupled, so load applied at one site produces deformation at others—an elementary consequence of continuum mechanics in a connected structure (Humphrey, 2002), and on the model above the mechanical distribution step.

The skin–fascia continuum constitutes an extensive collagen-rich compartment and is therefore a candidate site for distributed local transduction. Direct measurement of load-dependent electrical phenomena in dermis and superficial fascia under physiological conditions is sparse, and cutaneous electrical measurements in general admit alternative explanations including sweat-gland activity, hydration state and electrode artefact. The skin–fascia continuum is presented as a hypothesised site of interest and a practical target for measurement.

5.4 Relation to conventional mechanotransduction

What does the electrical component add to an account that already includes mechanotransduction?

Integrin–cytoskeletal signalling, mechanosensitive ion channels including PIEZO1, and YAP/TAZ- and TGF-β-dependent transcriptional programmes are established mediators of matrix-producing cell behaviour and fibrotic phenotype (Humphrey et al., 2014; Hinz, 2015; Zhao et al., 2022; Tomasek et al., 2002). Any adequate model must include them, and the present framework does not propose that SGPs replace them.

Two pathways from the same mechanical input are proposed, interacting rather than alternative.

5.4.1 Pathway A (established)

Mechanical deformation → cell-intrinsic mechanosensing via integrins, cytoskeleton and mechanosensitive channels, including transmission of force through the cytoskeleton to sites distant from the point of application (Wang et al., 2009) → transcriptional response → altered matrix production (Humphrey et al., 2014; Hinz, 2015; Zhao et al., 2022; Tomasek et al., 2002).

5.4.2 Pathway B (proposed)

Mechanical deformation of collagen-rich matrix → local strain-generated potential → modulation of membrane potential, ion flux and directional cues at the cell surface → proposed convergence with Pathway A at the level of calcium signalling and cytoskeletal organisation. The convergence node is proposed rather than demonstrated for endogenous SGPs; what is established is that in applied fields, fibroblast shape change and cytoskeletal reorganisation are calcium-dependent (Onuma and Hui, 1988), which makes calcium a plausible but unproven point of convergence.

Three properties would, if Pathway B operates at physiologically occurring magnitudes, make it non-redundant. Each requires qualification.

• It is a tissue property rather than a cell property. An SGP arises from deformation of the tissue rather than from cell-intrinsic sensing, and is therefore in principle present where cells are sparse or quiescent. This applies with different force to the different mechanisms: solid-state contributions (H2, H3) are properties of the solid matrix and would persist in matrix depleted of mobile fluid, whereas electrokinetic contributions require the fluid phase and mobile ions and are properties of the matrix–fluid system. H1 as a whole should therefore not be described as a property of the solid matrix alone.
• It may carry a sign. Reversal of the sense of deformation reverses the polarity of a piezoelectric potential, and reversal of flow direction reverses that of a streaming potential; polarity reversal has also been observed with changes in ionic composition (Pienkowski and Pollack, 1983). Whether tissue-level deformation maps onto a consistent field direction at the cell surface depends on the mechanism, the local geometry and the fibre architecture, and is not general. The claim here is that sign-carrying is possible in principle, not that a reliable vectorial code has been demonstrated.
• It is a candidate carrier of loading-rate information. Charge separation upon deformation is prompt relative to transcriptional responses. Whether this yields biologically distinguishable outcomes under matched mean load but different loading rate is untested and appears below as a prediction, not as a property.

The question the framework must answer is therefore precise: does the electrical component of a tissue’s response to deformation explain variance in cellular remodelling behaviour beyond that explained by mechanical strain itself?

6 The proposed causal chain

The model is stated as an explicit chain with each link graded, so that novel content is isolated from established background and from definitional content (Table 3).

Grading scheme . E – empirically established I – indirectly supported U – untested D – definitional within the proposed framework

TABLE 3. The proposed chain from load geometry to fibrotic stabilisation, with grading.

Links 1 (regionally), 2 (as modulation) and 6 are not claimed as contributions of this work. Link 7 is definitional. The proposal of this paper is links 3b, 3c, 4 and 5, and above all link 4. If link 4 fails, the framework reduces to a restatement of established mechanotransduction with additional terminology.

6.1 Integration with the FPA

If links 3–5 hold, an additional mechanism becomes available for how chronic load geometry may bias tissue fate: matrix deformation would supply electrical as well as mechanical information, arising wherever the load path produces deformation, and biased in direction wherever load geometry is chronically asymmetric.

Table 4 presents the corresponding staged progression as a hypothesis.

TABLE 4. Hypothesised progression from adaptive load distribution to fibrotic stabilisation. This is a proposed model. No longitudinal data establish these stages, their ordering, or their boundaries.

Fibrotic matrix differs from healthy matrix in collagen density, fibre orientation, crosslink density, water content and bulk electrical properties (Mir et al., 2025). Each could plausibly increase, decrease or redistribute strain-generated potentials. For a given applied force, higher stiffness yields smaller strain, tending to reduce deformation-coupled signal; altered fibre alignment and reduced water content act on the same quantity in the opposite direction, and reduced water content would attenuate the electrokinetic component, which requires fluid flow, while a solid-state component does not require flow and might therefore be relatively preserved—though hydration also affects collagen mechanics and dielectric properties, so its net effect on solid-state magnitude is itself unknown. The net effect is unknown and constitutes a measurable open question—one on which H2 bears directly.

7 Posture and ergonomics: a note on hierarchy

Ergonomic modification and postural training alter external load vectors and can reduce compensatory demand (McGill, 2007; van Dieën et al., 2019). We propose that they act as extrinsic modulators rather than determinants of internal load organisation: conventionally correct posture may coexist with high sustained compensatory activation, and the converse may also hold. On this view, observable posture is better treated as one output of load organisation than as its cause. This is a conjecture following from the framework, testable through the compensatory-activation measures indicated in §2.2.

8 Predictions and refutation criteria

Five predictions are given (Table 5). Each bears on H1, H2 or H3, and each is stated with the result that would refute or substantially weaken it, and with the experimental domain within which that refutation would apply.

TABLE 5. Predictions with measurement and refutation criteria.

A note on the scope of refutation. A negative result in a single material, cell type and loading regime constrains the hypothesis within that domain; it does not exclude every possible electromechanical mechanism in every tissue. Where “refuted” appears below, it means refuted for the domain tested, and the framework’s claim to generality diminishes accordingly with each such result.

8.1 On C2 and commensurability

A comparison of volts or volts per metre drawn from heterogeneous experiments is not sufficient and would be misleading. Applied-field studies and endogenous-potential measurements must be matched, or explicitly reconciled, on: field strength versus potential difference and the spatial scale over which each is defined; the dimension of the responding cell relative to that scale, since fibroblasts have been shown to respond to voltage gradients as small as a fraction of a millivolt across the cell width (Erickson and Nuccitelli, 1984); waveform, whether DC, cyclic or transient; duration of exposure and the duration required for the response in question; frequency content; and the geometry of the cell–matrix interface, since a potential generated within the matrix is not necessarily presented to the cell as a uniform extracellular field. Fibroblast thresholds themselves span roughly two orders of magnitude depending on which response is measured, the medium, and the exposure duration (Erickson and Nuccitelli, 1984); C2 must therefore specify which response is at issue before it can be evaluated.

8.2 On C5 and the asymmetry of temporal evidence

The two directions are not symmetrical, and the prediction is stated accordingly. If electromechanical change is consistently found to follow established matrix stiffening, this argues against a primarily instructive role and is consistent with SGP alteration being a consequence of altered matrix properties. If electromechanical change precedes histological change, this is consistent with an instructive role but does not establish causation, since both could follow a common upstream cause. C5 is therefore capable of weakening the framework more decisively than of confirming it, and confirmation would require the interventional design of C3.

8.3 Discriminating H2 from H3

C4 establishes whether a solid-state component exists but cannot distinguish piezoelectric from flexoelectric contributions, since both are solid-state and both would survive suppression of fluid movement. A further discriminating step is therefore required, and we state it separately rather than allowing C4 to be read as a test of piezoelectricity.

Two properties separate the mechanisms in principle, and neither yields a clean experiment.

• Scaling with strain gradient. Piezoelectric polarisation scales with strain; flexoelectric polarisation scales with the strain gradient and therefore grows as characteristic feature size falls. A systematic comparison across specimen thicknesses or feature sizes at matched mean strain would, in principle, separate a size-independent from a size-dependent contribution. This is a scaling argument, not a loading-mode argument: as set out in §5.1, bending does not isolate flexoelectricity and uniform extension does not isolate piezoelectricity, because a non-centrosymmetric and mechanically heterogeneous material generates both under any ordinary loading. Changing specimen geometry also changes the stress state, surface-to-volume ratio and fluid transport path, each of which is a confounder.
• Symmetry dependence. Piezoelectricity requires non-centrosymmetric molecular order; flexoelectricity does not. Preparations disrupting collagen’s polar order while preserving bulk mechanical properties and fluid content should abolish a piezoelectric contribution while leaving a flexoelectric one available. But disrupting molecular order also alters stiffness, fibre organisation and hydration behaviour, and the manipulation is therefore not specific.

A further constraint applies to the measurement technique itself. Because PFM generates an inhomogeneous field under the tip, it produces converse flexoelectric response in materials of any symmetry, and effective piezoelectric coefficients measured this way may not be piezoelectric in origin (Abdollahi et al., 2019); the reverse confusion has also been shown (Abdollahi et al., 2018). Nanoscale probe methods are accordingly poor arbiters of H3, whatever their value for H2.

We therefore state the H2/H3 discrimination as a required step and an unsolved one, and do not present H3 as testable by C4 or by any single loading configuration.

8.4 A candidate discriminating experimental strategy for C3

This is offered as a candidate strategy, not a decisive protocol.

8.4.1 Outline

Culture matrix-producing cells on two substrates—an electromechanically active collagenous or synthetic piezoelectric matrix, and an electrically inert matrix—apply an identical strain history, and assess α-SMA expression, collagen I and III synthesis, MMP/TIMP ratio, cytoskeletal organisation, YAP/TAZ nuclear localisation, and alignment relative to the strain axis.

The central methodological challenge is to isolate electrical output while preserving mechanically and chemically equivalent boundary conditions. Substrates differing in electromechanical activity typically differ also in bulk and local stiffness, surface topography, fibre organisation, porosity, surface chemistry, hydration behaviour and adhesion-ligand presentation, any of which could produce phenotypic differences independent of electrical output. Matching on bulk stiffness alone is insufficient, since cells respond to local as well as bulk mechanical properties. Development of adequately matched substrate pairs is a prerequisite for the experiment rather than a detail of its execution, and we do not claim that current materials meet this requirement.

8.4.2 Candidate control conditions, each requiring validation before it can carry argumentative weight

• Manipulation of medium conductivity. Increasing medium conductivity might attenuate an electrically mediated difference while leaving mechanical exposure unchanged. Whether it does so without altering osmotic conditions, cell surface charge, adhesion or matrix hydration, and whether the attenuation is quantitatively predictable, requires independent verification. This control is not established.
• Exogenous field application. Applying an external field to cells on the inert substrate is a weaker control than it first appears, because an externally imposed, spatially uniform field is not the same physical quantity as a potential generated within the matrix at the cell–matrix interface. The two differ in spatial distribution, in their relationship to the adhesion surface, and in their temporal coupling to the strain waveform. A positive result would show that fibroblasts in this system respond to fields of that magnitude; it would not show that they respond to the endogenous quantity. This control should be interpreted as establishing sufficiency of an applied field, not equivalence to the matrix-generated one.
• Manipulation of interstitial fluid movement, as in C4, and the strain–gradient and symmetry manipulations of §8.3.

8.4.3 Interpretation

If remodelling phenotype does not differ between substrates under adequately matched mechanical and chemical conditions, across the substrates, cell types and loading regimes tested, the electrical contribution proposed in links 3b–5 would not be supported within that domain, and the framework’s central claim would lose its principal support.

9 Limitations

The framework’s central weakness is stated in Table 3: link 4 is untested, and the framework stands or falls with it. It has not been shown that endogenous strain-generated potentials in non-osseous connective tissue reach magnitudes to which resident cells respond, and the two literatures bearing on the question have not been placed on a commensurable footing.

The tissue-level evidence for strain-generated potentials is predominantly osseous, whereas the framework’s principal claims concern fascia, dermis and other soft collagenous tissue. This extrapolation is the framework’s second major exposure.

The discriminating strategies of §8.3 and §8.4 are strategies rather than protocols. Substrate pairs matched on all mechanically and chemically relevant properties while differing in electrical output are demanding to construct; the proposed controls require validation; and the exogenous-field control tests a related but not identical physical quantity.

The step from regional biomechanics to an organism-level geometric architecture (Table 3, link 1) is an extrapolation this framework makes rather than one it inherits, and it has not been independently validated.

The terminology introduced here carries the risk common to new nomenclature: that naming a construct is mistaken for measuring it. Epistemic status is stated in Table 1 and no index formula is proposed, but the risk is not eliminated by these measures.

The framework does not exclude alternative or additional mechanisms. Biochemical signalling, inflammatory cascades, metabolic state and neural regulation operate alongside any electromechanical contribution and in most contexts are likely to dominate. The claim is not exclusivity but non-redundancy, and that claim is what requires testing.

H3 rests on a weaker evidential footing than H2. The nanoscale measurements most often cited as demonstrating collagen piezoelectricity were obtained by a technique that generates converse flexoelectric response in any dielectric, and they therefore bear more directly on H2 than on H3 (Abdollahi et al., 2019; Abdollahi et al., 2018). No currently available loading configuration or probe method isolates the piezoelectric from the flexoelectric contribution in collagenous tissue.

Finally, H2 and H3 may fail independently of H1 and of each other, and the experiments that would decide H3 are the least developed of those proposed.

10 Conclusion

We have proposed that chronic load geometry may bias tissue towards fibrotic stabilisation not only through established mechanotransductive pathways but through a distributed electromechanical component: deformation of collagen-rich matrix along a mechanically continuous load path generating local strain-generated potentials that are hypothesised to modulate the behaviour of matrix-producing cells. The proposal is structured as three nested hypotheses so that the unresolved status of collagen piezoelectricity under physiological hydration is carried as an explicit and separately testable claim.

Stated precisely, the proposed chain runs: load geometry → regional stress and strain → matrix deformation → local strain-generated potentials → modulation of matrix-producing cell behaviour → altered matrix architecture → altered reversibility. As a conceptual shorthand for that chain, and not as a physical equation, it may be summarised as pressure becomes electricity, electricity becomes instruction, and instruction becomes structure—with the caveat, set out in §1.1, that the operative quantity is deformation rather than hydrostatic pressure, and that whether each transition occurs at physiologically relevant magnitudes in living tissue is what remains to be shown.

Two measurements would do most to settle the question. The first is the magnitude of endogenously generated strain-generated potentials in non-osseous connective tissue under physiological loading, placed on a commensurable footing with fibroblast dose–response data. The second is the comparison of electromechanically active and inert matrices under adequately matched mechanical conditions. If the second shows no difference across the domains tested, the model proposed here loses its principal support.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. AI was used for language editing, structural organization, and refinement of scientific writing. All scientific concepts, interpretations, and conclusions were developed and verified by the author.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: George Maiti, New York University Health, United States

Reviewed by: Yong-wang Li, PLA Rocket Force Characteristic Medical Center, China

Pierre Claver Uzabakiriho, Wenzhou Medical University, China