Advanced quantitative imaging and biomechanical analyses of periosteal fibers in accelerated bone growth

Highlights

• Second Harmonic Generation imaging revealed differences in periosteal fibers at various strains.

• Measured the changes in periosteal growth factor expression at various strains

• Different regions of bone demonstrated complex regional differences in fiber orientation.

• Increasing strain had significant increase in relative mRNA expression for Ihh and PTHrP.

Abstract

Purpose

The accepted mechanism explaining the accelerated growth following periosteal resection is that the periosteum serves as a mechanical restraint to restrict physeal growth. To test the veracity of this mechanism we first utilized Second Harmonic Generation (SHG) imaging to measure differences of periosteal fiber alignment at various strains. Additionally, we measured changes in periosteal growth factor transcription. Next we utilized SHG imaging to assess the alignment of the periosteal fibers on the bone both before and after periosteal resection. Based on the currently accepted mechanism, we hypothesized that the periosteal fibers adjacent to the physis should be more aligned (under tension) during growth and become less aligned (more relaxed) following metaphyseal periosteal resection. In addition, we measured the changes in periosteal micro- and macro-scale mechanics.

Methods

30 seven-week old New Zealand White rabbits were sacrificed. The periosteum was imaged on the bone at five regions using SHG imaging. One centimeter periosteal resections were then performed at the proximal tibial metaphyses. The resected periosteal strips were stretched to different strains in a materials testing system (MTS), fixed, and imaged using SHG microscopy. Collagen fiber alignment at each strain was then determined computationally using CurveAlign. In addition, periosteal strips underwent biomechanical testing in both circumferential and axial directions to determine modulus, failure stress, and failure strain. Relative mRNA expression of growth factors: TGFβ-1, -2, -3, Ihh, PTHrP, Gli, and Patched were measured following loading of the periosteal strips at physiological strains in a bioreactor. The periosteum adjacent to the physis of six tibiae was imaged on the bone, before and after, metaphyseal periosteal resection, and fiber alignment was computed. One-way ANOVA statistics were performed on all data.

Results

Imaging of the periosteum at different regions of the bone demonstrated complex regional differences in fiber orientation. Increasing periosteal strain on the resected strips increased periosteal fiber alignment (p < 0.0001). The only exception to this pattern was the 10% strain on the tibial periosteum, which may indicate fiber rupture at this non-physiologic strain. Periosteal fiber alignment adjacent to the resection became less aligned while those adjacent to the physes remained relatively unchanged before and after periosteal resection. Increasing periosteal strain on the resected strips increased periosteal fiber alignment (p < 0.0001). The only exception to this pattern was the 10% strain on the tibial periosteum, which may indicate fiber rupture (and consequent retraction) at this non-physiologic strain. Increasing periosteal strain revealed a significant increase in relative mRNA expression for Ihh, PTHrP, Gli, and Patched, respectively.

Conclusion

Periosteal fibers adjacent to the growth plate do not appear under tension in the growing limb, and the alignments of these fibers remain unchanged following periosteal resection.

Significance

The results of this study call into question the long-accepted role of the periosteum acting as a simple mechanical tether restricting growth at the physis.

Introduction

Previous studies have shown that growth acceleration can occur in many animal species after periosteal procedures [1], [2], [3], [4], [5], and periosteal resection (removal of a strip of periosteum) has been shown to accelerate physeal growth with minimum morbidity [6], [7], [8], [9], [10], [11], [12]. Recently, we have shown that circumferential periosteal resection reproducibly accelerates growth in a rabbit model [13]. The accepted mechanism for this growth acceleration is that the periosteum serves as a mechanical tether to restrict physeal growth [14], [15]. If true, one would expect: (1) periosteal fibers adjacent to the physis to be under tension during growth and (2) periosteal fibers adjacent to the physis to relax following metaphyseal periosteal resection.

In addition to periosteal fibers mechanical properties and structure, periosteal cells have been shown to express proteins capable of regulating growth at the physis [16], [17]. Many of these factors are involved in complex signaling pathways that overlap one another. Parathyroid hormone-related protein (PTHrP) and Indian Hedgehog (Ihh) [18] are such factors, important to endochondral ossification and prenatal growth; however, there is a relatively large gap in knowledge of the regulation of these two paracrine factors, especially regarding their role in post-natal growth. Ihh has the ability to regulate Patched and Gli. Both of these factors are conserved targets of Hedgehog signaling and are required for a cellular response in this signaling pathway [19]. Similarly, Transforming Growth Factors Beta (TGFβ (1–3)) has been shown to play a critical role in bone remodeling, promote osteoprogenitor proliferation and differentiation [20], and may interact with cells in the perichondrium to affect the PTHrP-Ihh feedback loop [21]. Moreover, recent studies have shown that TGFβ knock outs are prone to limb deformity such as reduced bone growth, expansion, and mineralization [22], [23]. In addition to the expression of these factors in the periosteum, their expression appears to also be regulated by mechanical stimuli. These results are similar to those previously described investigating PTHrP expression near entheses [16], in which PTHrP expression was found to be elevated when tendons and ligaments were loaded and diminished upon removal of the mechanical stimuli. Therefore, it is plausible that changes in periosteal fiber morphology and alignment during normal growth and following periosteal resection may result in different mechanical stimuli, which may provide differential expression of growth factors in the periosteum. This relationship may help explain the association between the periosteum and physeal growth.

While removing the periosteum from the bone allows one to characterize the periosteum's mechanical and biologic properties, the alignment and packing of periosteal fibers has not been well characterized, especially on a maturing animal [24], [25]. Foolen et al. were the first to successfully demonstrate periosteal fiber organization in chick embryo periosteum using two-photon excitation fluorescence (TPEF) microscopy via exogenous stains, but this was performed with the bone removed at a developmental stage well-before full maturity [26], [27]. Second Harmonic Generation (SHG) imaging microscopy may offer a superior tool in quantifying the changes in collagen distribution and orientation of tibial periosteum than TPEF since SHG can acquire three-dimensional, sub-micron resolution images of non-centrosymmetric assemblies, such as collagen fibers in connective tissue extracellular matrix (ECM) without the use of exogenous contrast agents [28], [29], [30], [31], [32], [33]. This coherent process occurs when two photons simultaneously combine to form a second order polarization in a tissue and emit a photon of exactly twice the frequency governed by:

$$\overrightarrow{\mathrm{P}}={\chi }^{\left(2\right)}\overrightarrow{\mathrm{E}}\cdot \overrightarrow{\mathrm{E}},$$

where P is the induced polarization; χ⁽²⁾ is the second-order nonlinear susceptibility tensor (bulk property measured in the experiment); and E is the electric field vector of the laser [34]. The second-order susceptibility can be described by the following:

$${\chi }^{\left(2\right)}={\mathrm{N}}_{\mathrm{S}}<\mathrm{\beta }>,$$

where N_s = density of molecules, and β = first molecular hyperpolarizability (brackets denote their orientational average). SHG contrast is formed due to the molecular-level property of nonlinearity (β), and this contrast mechanism is defined in the terms of a permanent dipole moment (d):

$$d^2=\mathrm{\beta EE},$$

where E = the electric field due to a dipole. The relative concentration and alignment of the dipoles are reflected by the magnitude of χ⁽²⁾, as SHG contrast increases for well-ordered structures due to the second order symmetry constraints imposed by χ⁽²⁾[33]. Due to specific phase-matching considerations of the SHG process, sub-fiber resolution can be acquired as the emission direction contains information related to the fibril architecture, unlike the spatially isotropic emission of TPEF [35], [36], [37], [38].

Currently, there is a lack of characterization regarding imaging, biomechanics, and mechanobiology of the mammalian periosteum in the literature [39]. In this study we combine multiple analyses of the periosteal strips removed during periosteal resection and SHG imaging of the periosteum remaining on the bone, in an effort to confirm or refute the accepted “un-tethering” mechanism of growth acceleration following periosteal resection. Using the explanted metaphyseal perisosteal strips, we: (1) utilized SHG imaging to measure changes in periosteal fiber alignment in response to various mechanical strains; (2) biomechanically characterized the metaphyseal periosteal strips; (3) calculated the periosteal stress necessary to restrict growth and compared these values with the values found from mechanical testing; and (4) evaluated relative mRNA expression of paracrine signaling factors from periosteal cells under mechanical stimulation. In addition, we utilized SHG imaging to assess, for the first time reported, the alignment of the periosteal fibers on the bone before and after periosteal resection.