What Happens After the Injection?
We have become very used to talking about aesthetic injectables in terms of what they are supposed to do.
This one hydrates. That one stimulates fibroblasts. Another promotes collagen. A biostimulator remodels the extracellular matrix.
It is useful shorthand, but it also makes the biology sound unusually straightforward. Put something into the skin, the fibroblasts receive a signal, collagen production increases, and the tissue gradually improves.
The reality is rather more complicated. An injectable enters living tissue with its own architecture, mechanical properties, resident cells, immune cells and continually changing extracellular matrix. The material then becomes part of that environment, and the response depends on the interaction between the two.
So I became curious about a slightly different question: what does the tissue actually experience after we put something there?
The skin is not a blank canvas
The dermis is a connective tissue in which fibroblasts sit within a three-dimensional extracellular matrix made largely from collagen and other structural molecules. The matrix provides much of the skin's physical structure, while also influencing how the cells within it behave.
That relationship runs in both directions. Fibroblasts produce and remodel the extracellular matrix, while the matrix influences fibroblast activity in return. Collagen is continually being synthesised, organised and broken down, with matrix metalloproteinases and their inhibitors helping regulate that process.
The extracellular matrix therefore has a role in signalling as well as structure. Frantz et al. (2010), Journal of Cell Science describe the ECM as providing biochemical and biomechanical cues that influence tissue homeostasis, cell behaviour and remodelling.
Once you look at skin this way, an injectable becomes part of a larger biological system. What matters is the material itself, the environment it enters and the changes that occur around it.
Cells can feel their surroundings
Fibroblasts are mechanically connected to the extracellular matrix through structures involving integrins, focal adhesions and the actin cytoskeleton. These connections provide physical attachment, but they also transmit information about forces within the surrounding tissue.
This is part of what is known as mechanotransduction. A physical change in the cellular environment can be converted into biochemical signalling inside the cell, influencing processes such as gene expression and matrix production. Sun et al. (2016), The Journal of Cell Biology reviews the role of integrins in this process, including their connections to the cytoskeleton and their ability to respond to mechanical forces.
That has an obvious relevance to injectable materials. A fibroblast does not necessarily need to bind directly to a product for the product to influence it. A change in the forces around the cell may alter its behaviour.
Hyaluronic acid gives us a particularly clear human example.
HA: when the physical environment becomes part of the response
In a 2007 study by Wang and colleagues, Archives of Dermatology, cross-linked hyaluronic acid was injected into photodamaged forearm skin in 11 older volunteers, with saline used at control sites. Biopsies were taken four and thirteen weeks later, allowing the researchers to examine the tissue response directly.
They found increased markers of newly synthesised type I collagen around the filler, together with increased type I and III procollagen gene expression. The fibroblasts surrounding the filler also had a stretched, biosynthetically active appearance. When the researchers examined the interaction in vitro, they did not find direct binding between the fibroblasts and the filler, leading them to propose that mechanical stretching of the surrounding dermis could be involved in the response.
That interpretation was explored further in a 2013 randomised, placebo-controlled study by Turlier and colleagues, Journal of Dermatological Science. Sixty women received HA or saline in the arms, with biopsies collected at baseline and after one, three and six months. The researchers measured procollagen, collagen-related gene expression, MMPs and TIMP-1, while also using ultrasound to follow the filler itself.
At one month, dermal procollagen was 66% higher on the filler-treated side than the control side. The filler volume then declined over time, with approximately 43% of the original volume remaining at one month, 26% at three months and 20% at six months. The biochemical response also diminished as the material dispersed.
The authors linked these findings to the mechanical effects of the filler, while acknowledging that the precise cellular signalling mechanism had not been established.
A 2024 human study by Wang and colleagues, Experimental Dermatology followed the same line of investigation with biochemical and microscopic analyses. Fibroblast activation was observed from one week after injection and continued for several months, while microscopy showed elongation of fibroblasts at one week and thick collagen bundles accumulating around pools of cross-linked HA by four weeks. Substantial injected HA and accumulated collagen were still present at twelve months.
Taken together, these studies make a fairly coherent case for the mechanical environment being part of the response to cross-linked HA in photoaged human skin. The precise contribution of mechanical signalling, however, is still a mechanistic question rather than something we can simply assume is identical across every HA product.
That seems worth remembering whenever "HA stimulates collagen" appears on a treatment menu.
PLLA: when a material becomes part of the tissue response
Poly-L-lactic acid creates a different situation because it is a particulate biodegradable polymer.
In a 2013 human study by Goldberg and colleagues, Dermatologic Surgery, 14 healthy volunteers received injectable PLLA and underwent biopsies at three, six and twelve months. Type I collagen increased significantly at three and six months, while type III collagen also increased numerically without reaching statistical significance.
The twelve-month analysis needs some caution because the samples were affected by technical differences in the collagen staining procedure. The strongest quantitative evidence in that study therefore comes from the earlier time points.
A more detailed picture comes from a 2015 study by Stein and colleagues, Journal of Dermatological Science, which examined tissue surrounding PLLA particles in treated human skin.
Macrophages were found immediately around detectable PLLA particles, with fibroblasts in a surrounding layer. Alpha-SMA-positive myofibroblasts and vascular structures were also observed. The distribution of collagen was particularly notable: type III collagen was concentrated close to the PLLA particles, while type I collagen appeared towards the periphery of the surrounding encapsulation.
The investigators also found increased expression of collagen I and III, TGF-β1 and TIMP1. They interpreted the findings as part of a foreign-body response in which macrophages, fibroblasts and myofibroblasts participate in the gradual formation of tissue around the particles.
Animal work has taken that question further. In a 2023 study in Cells, Oh and colleagues examined PLLA in aged animal skin and found changes involving macrophage polarisation, fibroblast activity and collagen synthesis. It provides useful mechanistic evidence, although it is still animal research and should not be treated as a direct demonstration of the same pathway in human skin.
PLLA therefore gives us a different way to think about collagen production. The particle remains within the tissue while the surrounding cells respond to it, and the eventual matrix is part of that response.
PDRN: when the material may change along the way
PDRN introduces yet another possibility.
In a 1999 study by Thellung and colleagues, Life Sciences, primary human skin fibroblasts from seven donors were exposed to PDRN. The researchers found increased fibroblast proliferation and then used receptor agonists and antagonists to investigate the mechanism.
The response was reduced by an A2 adenosine receptor antagonist and reproduced by an A2 agonist. PDRN also produced an increase in intracellular calcium, supporting a role for A2-purinergic signalling.
The authors proposed that the DNA polymer itself might not be the complete story. They suggested that PDRN could be broken down into smaller nucleotides, nucleosides and bases that could then participate in receptor-mediated signalling. The preparation used in the experiment consisted predominantly of DNA polymers with chain lengths of approximately 50 to 2000 base pairs.
There was another curious result. Intact PDRN produced the proliferative response, while the acute calcium response required sonication. The authors therefore considered whether smaller degradation products were involved in that part of the response.
This is useful mechanistic evidence, but it has to stay within its limits. It was a fibroblast study using a specific PDRN preparation, rather than a biopsy study of a modern PN injectable in living human skin. It gives us a possible biological pathway; it does not establish that every PN or PDRN product works through the same mechanism.
That distinction becomes particularly important because PN and PDRN products can differ in molecular characteristics, formulation and processing.
More collagen does not tell us everything
Collagen is an important part of the story, but it isn't the whole tissue.
Its type matters. Its organisation matters. Where it is deposited matters. The balance between synthesis and degradation matters as well. So does the behaviour of the cells producing it.
This is especially relevant when we talk about regeneration. Adult human skin is very effective at repairing injury, but repair can involve fibrosis rather than a complete recreation of the original tissue architecture. Dermal fibroblasts are heterogeneous, and their behaviour can contribute to either productive remodelling or scarring depending on the signals and matrix around them.
The distinction between repair and regeneration is therefore more useful than simply counting collagen.
A treatment can increase collagen without telling us exactly what kind of tissue has been produced or how that tissue is organised.
That leaves a much more interesting set of questions. Where did the material go? How long did it remain? Which cells interacted with it? What happened to the extracellular matrix around it? Did the mechanical environment change? Did immune cells participate? What happened as the material degraded?
Those details help explain why two treatments that are both described as "collagen stimulators" can have very different biological behaviour.
So what are we actually injecting?
The more I look at this literature, the more useful the physical side of the story becomes.
An injectable has an ingredient list, but it also has a size, structure, viscosity, elasticity, persistence and pattern of distribution. Once it enters tissue, those properties determine what cells and matrix encounter. The surrounding tissue then responds, and that response can change over time as the material disperses, degrades or becomes surrounded by newly deposited matrix.
With cross-linked HA, human studies suggest that changes in mechanical support and fibroblast behaviour are closely linked to the subsequent collagen response. With PLLA, the material becomes associated with a cellular response involving macrophages, fibroblasts, myofibroblasts and collagen deposition around the particles. With PDRN, cellular studies suggest that purinergic signalling and the processing of DNA-derived material may contribute to the response.
These are different biological situations, even though they can all end up being described with the same phrase: biostimulation.
Perhaps that is the more useful way to understand the term. It describes an intended effect rather than a single mechanism. The material matters, but so do the cells encountering it, the extracellular matrix around them, the immune response and the condition of the tissue before treatment.
Which brings me back to the question I started with: what actually happens after we put something into skin?
The answer seems to be considerably more complicated than "it stimulates collagen". An injectable may have a defined composition, but once it enters living tissue, what happens depends on the conversation between that material and the tissue around it.
And that is probably a more useful place to start when trying to understand what these treatments are actually doing.
Research notes
This piece draws on histology, extracellular-matrix biology, mechanotransduction, wound healing and injectable biomaterials. The human evidence is strongest for the HA and PLLA sections. The PDRN discussion relies more heavily on cellular and mechanistic evidence, so the conclusions there are deliberately cautious.
References
Mescher AL. Junqueira's Basic Histology: Text and Atlas. 16th ed. McGraw Hill; 2021.
Alberts B, Heald R, Johnson A, et al. Molecular Biology of the Cell. 7th ed. Garland Science; 2022.
Frantz et al. (2010), Journal of Cell Science. The extracellular matrix at a glance. 123(24):4195-4200. doi:10.1242/jcs.023820.
Sun et al. (2016), The Journal of Cell Biology. Integrin-mediated mechanotransduction. 215(4):445-456. doi:10.1083/jcb.201609037.
Jiang and Rinkevich (2020), International Journal of Molecular Sciences. Scars or regeneration? Dermal fibroblasts as drivers of diverse skin wound responses. 21(2):617. doi:10.3390/ijms21020617.
Wang et al. (2007), Archives of Dermatology. In vivo stimulation of de novo collagen production caused by cross-linked hyaluronic acid dermal filler injections in photodamaged human skin. 143(2):155-163. doi:10.1001/archderm.143.2.155.
Turlier et al. (2013), Journal of Dermatological Science. Association between collagen production and mechanical stretching in dermal extracellular matrix: In vivo effect of cross-linked hyaluronic acid filler. A randomised, placebo-controlled study. 69(3):187-194. doi:10.1016/j.jdermsci.2012.12.006.
Wang et al. (2024), Experimental Dermatology. Implications for cumulative and prolonged clinical improvement induced by cross-linked hyaluronic acid: An in vivo biochemical/microscopic study in humans. 33(1):e14998. doi:10.1111/exd.14998.
Goldberg et al. (2013), Dermatologic Surgery. Single-arm study for the characterization of human tissue response to injectable poly-L-lactic acid. 39(6):915-922. doi:10.1111/dsu.12164.
Stein et al. (2015), Journal of Dermatological Science. The biological basis for poly-L-lactic acid-induced augmentation. 78(1):26-33. doi:10.1016/j.jdermsci.2015.01.012.
Oh et al. (2023), Cells. Poly-L-lactic acid fillers improved dermal collagen synthesis by modulating M2 macrophage polarization in aged animal skin. 12(9):1320. doi:10.3390/cells12091320.
Thellung et al. (1999), Life Sciences. Polydeoxyribonucleotides enhance the proliferation of human skin fibroblasts: involvement of A2 purinergic receptor subtypes. 64(18):1661-1674. doi:10.1016/S0024-3205(99)00104-6.