Imaging technologies are essential for advancing our understanding of biology, because scientific insight depends on what we can observe and measure. Zacharias Janssen’s invention of the microscope in the early seventeenth century revolutionized biology by revealing microorganisms and establishing the foundations of cell biology. Over the centuries, microscopy has progressed from brightfield and widefield techniques to confocal systems, which introduced optical sectioning and improved resolution. Today, nonlinear fluorescence microscopy represents the next leap forward, enabling deeper, more precise imaging of cells in living tissues.

Nonlinear fluorescence methods, such as two-photon and three-photon microscopy, harness the near-simultaneous absorption of multiple low-energy photons to generate highly localized excitation at the focal plane [1, 2]. By confining excitation to the focal volume, these approaches deliver reduced phototoxicity, minimal out-of-focus background, and superior penetration depth compared to earlier techniques. Moreover, innovations in laser technology, adaptive optics, and genetically encoded fluorescent reporters have enhanced imaging speed, signal fidelity, and molecular specificity, establishing intravital imaging as an indispensable tool for studying cellular physiology in situ.

Bone mechanobiology has particularly benefited from this technological evolution. The dense, mineralized collagen matrix of bone—which once posed a formidable barrier to light penetration—can now be traversed to visualize osteocytes in their native environment (Fig. 1). Two-photon excitation provides the depth reach and resolution necessary to track osteocyte calcium signaling and cytoskeletal dynamics, while three-photon modalities and adaptive optics extend visualization even deeper. Coupled with specialized fluorescent probes and surgical preparations, these advances enable real-time observation of mechanotransduction pathways within living bone, opening new avenues for understanding diseases like osteoporosis [3,4,5].

Fig. 1

Scanning electron microscopy images of osteocytes (A) embedded in bone matrix and (B) acid-etched, revealing processes emanating from the cell body. (Image B taken by Prof. Lynda Bonewald)

In this review, we provide a focused overview of the development and refinement of intravital imaging tools and techniques as applied to osteocyte research. We synthesize key discoveries from pivotal studies that have leveraged these approaches to uncover novel insights into bone-cell behavior, signaling, and function within the living bone environment.

Bioreactor Bone Core Experiments

Weinbaum and Cowin developed mathematical models predicting that fluid forces similar to those that activate cells in vitro would activate osteocytes in situ[6]. Subsequent in vitro fluid-flow studies demonstrated that osteocytes respond to fluid shear by secreting cytokines that regulate osteoclast and osteoblast activity [7,8,9,10]. Before the advent of high-resolution intravital imaging, researchers sought ways to preserve osteocytes in their native tissue while applying controlled mechanical stimuli to further interrogate this phenomenon. One of the earliest approaches employed bioreactor-based bone-core explants: small cylinders of living bone maintained viable ex vivo for experimental loading.

Initial work with canine trabecular cores showed that cyclic mechanical loading triggered osteocytic release of prostaglandins and prostacyclins, providing biochemical evidence of mechanotransduction within intact bone [11]. Subsequent refinements extended loading durations to probe sustained responses, revealing time-dependent patterns of cytokine production and gene expression in bone cores under repeated strain [12]. The field progressed further with the introduction of 3D bioreactor systems using human trabecular bone. These platforms enabled simultaneous measurement of dynamic markers—including lactate dehydrogenase activity, mineral apposition rate, and bone formation rate—under precisely controlled load regimens [13]. Although these studies preserved tissue architecture and yielded valuable bulk readouts, they relied primarily on endpoint biochemical assays rather than real-time cellular imaging.

Together, these bioreactor experiments established proof of principle that osteocytes within their native matrix sense and transduce mechanical cues. They also underscored the need for tools capable of capturing immediate, localized cellular responses—paving the way for modern intravital imaging strategies.

Real-Time Osteocyte Imaging Ex Vivo

Early visualization of the lacuno-canalicular network (LCN) in bone explants employed confocal microscopy of pre-stained samples. Static three-dimensional reconstructions used fluorescein [14] and procion red [15] to map mineralized canaliculi within sacrificed tissue. The introduction of two-photon microscopy enabled live imaging of the LCN in situ. Tracers such as calcein [16] and rhodamine B [17, 18] provided optical sectioning deep within intact explants. Fluorescent phalloidin staining delineated F-actin networks, revealing cytoskeletal organization during mechanostimulation [19, 20]. Moreover, cutting-edge multiplex protocols have allowed simultaneous visualization of structural and signaling markers, refining spatial and temporal resolution of osteocyte responses (Fig. 2) [21].

Fig. 2

Confocal microscopy provides high-resolution, static three-dimensional views of osteocyte structure and morphology; however, its ex vivo nature limits the ability to capture dynamic cellular behaviors, thereby inspiring the development of intravital imaging for real-time in vivo studies (A) Confocal image of the osteocyte lacunocanalicular system in a FITC stained, undecalcified mouse femur. Bar = 10 μm. A2 shows an enlarged image of the boxed area in A1 and A3 shows a 3D rendered view of the same osteocyte lacuna. The lower panel depicts widefield fluorescent images of whole mount neonatal mouse calvarium. (B) Immunostaining for E11/gp38 (cy3 anti-hamster antibody [red]) was combined with (C) Alexa Fluor 488-phalloidin staining for F-actin [green]. (D) The merged images highlight the capabilities of multiplexed confocal fluorescent imaging for evaluating osteocyte cell morphology in situ. Bar = 10 μm. These high-resolution images, made possible by confocal technology, are crucial for studying the complex 3D architecture where osteocytes reside and communicate, thus shedding light on their mechanical and physiological roles in bone. A.

Adapted from 10.1016/j.bone.2019.01.025 with permission [70]. B-D. Adapted from 10.1016/j.bone.2015.02.011 with permission [71]

Synthetic fluorophores, however, faced inherent limitations in dense bone: light scattering curtailed penetration, photobleaching reduced imaging duration, non-specific binding diminished signal clarity, and rapid dye clearance restricted longitudinal studies [22]. To reduce optical scatter, researchers adopted thin embryonic chick calvariae preparations (Fig. 3A). Ishihara et al. optimized this model for high-speed calcium imaging, minimizing photobleaching and resolving synchronized intercellular Ca2⁺ waves mediated by gap junctions and endoplasmic reticulum stores [23]. Tanaka et al. further integrated digital image correlation with 3D time-lapse Ca2⁺ microscopy on mechanically stimulated calvariae, correlating local strain fields with osteocyte morphology and signaling maturation (Fig. 3B). Functional imaging approaches in long bone soon followed (Figs. 3C & D) [24]. Calcium-sensitive dyes (e.g., Fluo-8 AM) captured real-time intracellular Ca2⁺ transients in osteocytes under mechanical or chemical stimulation (Fig. 4) [25]. This work validated in vitro results [26] suggesting that osteocytes were more mechanosensitive than osteoblasts, thus further supporting their role as bone mechanotransducers [27]. In work from the same group, Lifeact transgenic mice and Fluo-8 AM in ex vivo tibia compression experiments showed that Ca2⁺ oscillations precede actomyosin contractions, supporting an interesting model of calcium-dependent contractile mechanotransduction [28].

Fig. 3

Development of imaging strategies for visualizing osteocyte responses to mechanical loading across ex vivo, in situ, and in vivo systems. (A) Adachi et al. [72] established an ex vivo imaging platform using embryonic chicken calvaria bone fragments mounted with a 2 mm clearance, enabling direct microneedle-induced mechanical deformation during live imaging. (B) Tanaka et al. [24] implemented a fluid-flow setup for 3D time-lapse calcium imaging of osteocytes and osteoblasts within intact embryonic chick calvariae. (C) Jing et al. [25] designed an ex vivo mechanical loading device for in situ calcium and FRAP (Fluorescence Recovery After Photobleaching) imaging in mouse tibiae, utilizing a piezoelectric linear actuator to apply axial cyclic compression for precise mechanical stimulation during imaging. (D) Lewis et al. [33] developed a custom in vivo loading and imaging apparatus to apply controlled physiological deformation to mouse third metatarsal (MT3) bones while simultaneously capturing intracellular calcium responses via two-photon microscopy. A. Adapted from 10.1016/j.jbiomech.2009.07.006 with permission [72]. B. Adapted from 10.1007/s00774-017–0868-x with permission [24]. C. Adapted from 10.1096/fj.13–237578 with permission [25]

Fig. 4

Fluorescently labeled Ca.2+ was imaged in tibial explants with and without physiologically relevant mechanical loading. Autonomous and mechanical loading-induced Ca2 + responses in osteocytes of intact murine tibiae under either static or cyclic mechanical loading with an 8-N peak load magnitude. Figure Adapted from 10.1096/fj.13–237578 with permission [25]

These experiments offered previously unprecedented access to osteocytes in situ and validated in vitro and theoretical concepts about osteocyte function. They represent foundational steps toward building in vivo investigations of osteocyte mechanobiology and remain critical tools for studying bone biology (Fig. 3). While ex vivo studies preserve native tissue architecture and revealed immediate osteocyte responses, they often remain constrained by the drawbacks of synthetic dyes—limiting imaging depth, duration, and repeatability. These challenges have driven the development of approaches that leverage genetically encoded reporters and biosensors, which offer improved specificity, photostability, and compatibility with in vivo intravital imaging of osteocyte function.

In Vivo Interrogation of Osteocyte Mechanobiology

For decades, the deep embedding of osteocytes within the mineralized bone matrix posed a significant challenge for observing their real-time behavior in living animals. This began to shift with the advent of transgenic tools. A key milestone was the development of the Dmp1-Cre mouse line, which enabled specific genetic manipulation of osteocytes in vivo [29]. This system has been a key enabling technology for producing foundational knowledge about osteocyte function, establishing with certainty that osteocytes are not passive placeholders but active regulators of bone remodeling and mineral homeostasis [30,31,32]

The combination of this targeting strategy with fluorescent reporters enabled dynamic visualization of osteocyte activity. Transgenic mice expressing genetically encoded calcium indicators (GECIs) in an osteocyte targeted (i.e., DMP1-Cre dependent) manner allowed real-time imaging of calcium signaling during mechanical loading [33,34,35]. Multiphoton microscopy was used to capture osteocyte signaling in the third metatarsal during simultaneous tissue-level loading while maintaining primary blood and endocrine interactions intact. These experiments provided direct evidence of osteocyte responsiveness to physiologic strain and demonstrated that cortical bone osteocytes respond differentially based on strain magnitude, with broader cellular activation observed at higher strain levels. These findings align with long-standing in vivo data showing that bone formation rates scale with strain magnitude [36].

Subsequent studies using this system revealed that osteocyte mechanosensitivity is directly modulated by circulating hormones. In an ovariectomy model, estrogen depletion attenuated the osteocyte response to loading, providing the first in vivo evidence that endocrine signaling influences osteocyte mechanotransduction [37, 38]. Recent work has further shown that acetylcholine receptor signaling plays a role in load-induced osteocyte activation in a sex- and strain magnitude–dependent manner [39]. Because this model maintains both local and systemic signaling, it offers a valuable platform to study osteocyte-intrinsic pathways and tissue crosstalk mechanisms.

Together, these advances have added to the growing body of literature working to reshape our view of osteocytes—from passive matrix-bound cells to active sensors and integrators of mechanical and metabolic signals. Nonetheless, there is still room for technical advancement to expand investigative capabilities. Continued innovation is needed to develop specific, scalable, and broadly accessible tools for osteocyte imaging, particularly in deep tissues and pathologic contexts.

Expanding In Vivo Imaging to Subcellular Osteocyte Biology

While GECIs are well-suited for targeted fluorescent expression in vivo, the Dmp1-Cre system is associated with off-target expression and requires complex breeding strategies [35, 40, 41]. To complement genetic tools, researchers have developed fluorescent in vivo imaging approaches that circumvent these limitations. Intravital multiphoton microscopy has been used to visualize auto fluorescent bone signals, including collagen [42] and NADH [43]. Yuan et al. used this technique to image osteocytes and associated vasculature in cranial bone with injectable tracers [44]. They also employed label-free NADH autofluorescence to assess osteocyte metabolism during ischemic challenge, directly linking cellular bioenergetics to pathophysiological stress.

Normal osteocyte function depends on the unique three-dimensional cellular microenvironment of the lacuno-canalicular system. The refinement of intravital imaging techniques has made it possible to interrogate basic osteocyte functions in vivo while preserving local and systemic signaling inputs. A key opportunity lies in visualizing subcellular processes that were previously inaccessible in living bone. Traditional intravital approaches often relied on genetically encoded probes, which required complex and time-consuming breeding strategies, or employed conventional fluorophores limited by insufficient brightness and light scattering in mineralized tissues [33, 45, 46]. As a result, in vivo imaging was generally constrained to cytosolic phenomena such as calcium signaling, with limited resolution of finer intracellular dynamics.

To address these limitations, Matthews et al. introduced an innovative strategy using Cornell Prime Dots (C’Dots), ultra-bright, nanoparticle-based fluorescent probes [41]. C’Dots offer advantages including rapid cellular uptake, compatibility with local injection, and enhanced tissue penetration, increasing the flexibility of experimental design (Fig. 5). Most notably, C’Dots have enabled visualization of subcellular events—such as endocytosis and membrane remodeling—in osteocytes within live bone tissue. These studies complement and extend emerging investigations into the role of extracellular vesicles in osteocyte mechanobiology offering a multi-scale view of how subcellular events contribute to tissue-level signaling [47]. C’Dots can be functionalized as ratiometric probes [48,49,50], and thus have extensive experimental potential for intravital imaging of subcellular biochemical events. This approach expands the utility of intravital microscopy beyond mechanotransduction, positioning it as a tool for probing fundamental aspects of osteocyte cell biology in vivo with unprecedented resolution and specificity.

Fig. 5

The use of C'Dots nanoparticles (PEG-C'Dots and RGD-C'Dots) significantly enhances the capabilities of high-resolution multiphoton intravital imaging, enabling detailed visualization of osteocyte cell bodies, dendritic networks, and subcellular localization within live mouse metatarsal bone A-F) High resolution multiphoton intravital images of C’Dots within cortical osteocytes in a mouse metatarsal bone. Images were acquired with a 5 min incubation of C’Dots prior to imaging. Frame averaging and PMT gain were increased to improve structural resolution. PEG-C’Dots: A) 40 × zoom of a mouse MT3 allows visualization of edges of the bone, as well as dendrites spreading from osteocyte cell bodies. B) Subcellular localization of C’Dots is visible (purple inset, purple arrows). C) Dendritic networks between osteocytes can also be observed (yellow inset, yellow arrows). RGD-C’Dots: D) 80 × zoom of a mouse 3MT allows visualization of edges of the bone, as well as dendrites spreading from osteocyte cell bodies. E) Bright spots along osteocyte dendrites are visible (red inset, red arrows). F) Subcellular localization of RGD-C’Dots within the cell body (green inset, green arrows).

Adapted from 10.1016/j.bone.2023.116830 with permission [41]

Intravital Imaging of Other Bone Cell Types

While significant efforts have focused on overcoming the challenges of visualizing osteocytes, the advanced capabilities of intravital microscopy have also been instrumental in revealing the dynamic behavior and interactions of other key cell populations within the bone microenvironment. In particular, intravital imaging has become a critical tool for dissecting the real-time cellular dynamics that drive bone remodeling.

Studies by McDonald et al. and others have directly visualized the interactions between mature osteoblasts (mOBs) and osteoclasts (mOCs), revealing that direct mOB–mOC contact suppresses osteoclastic bone resorption [51]. These findings highlight the active regulatory role of osteoblasts in modulating bone turnover in situ [51, 52]. In addition, functional intravital imaging with pH-sensitive probes and proton pump reporter mice has enabled real-time visualization of acidification at osteoclast sealing zones. These tools have revealed marked functional heterogeneity among osteoclasts, distinguishing highly active, non-motile resorbing cells from less active, motile ones, and have facilitated rapid assessment of osteoclast responses to antiresorptive therapies such as bisphosphonates [45, 46, 53].

Beyond bone-resorbing and bone-forming cells, intravital microscopy has illuminated dynamic processes within the bone marrow that are difficult to capture in static or ex vivo analyses. This includes immune cell trafficking [54, 55], stromal interactions[55,56,57], and vascular remodeling (Fig. 6) [