Scalable production of controllable dermal papilla spheroids on PVA surfaces and the effects of spheroid size on hair follicle regeneration

Abstract

Organ size and numbers are vital issues in bioengineering for hair follicle (HF) regeneration. Murine HF dermal papilla (DP) cells are able to induce HF neogenesis when transplanted as aggregates. However, how the preparation of murine and human DP aggregates affects HF inductivity and the size of regenerated HF is yet to be determined. Here we report a scalable method for production of controllable human and rat DP spheroids in general labs for reproducible experiments. Compared with more hydrophobic polyethylene and poly(ethylene-co-vinyl alcohol), DP cells are poorly adhesive to hydrophilic polyvinyl alcohol (PVA). Seeded in PVA-coated 96-welled commercial PCR tube arrays, DP cells quickly aggregate into single spheroids with progressive compaction. Varying seeded cell numbers and culture periods enables us to control the size and cell number of the spheroids. The spheroids obtained have high viability and preserve DP characters. A proof of principle experiment was conducted to examine the size effect on the efficiency and efficacy of HF regeneration. We found that both human and rat DP spheroids are able to induce HF neogenesis and larger DP spheroids exhibit higher HF inductivity. However, the average diameter of regenerated hair fiber did not significantly change with the increasing size of transplanted DP spheroids. The result suggests that an appropriate size of DP spheroid is essential for HF inductivity, but its size cannot be directly translated to a thicker regenerated hair. Our results also have implications on the efficiency and efficacy in the regeneration of other epithelial organs.

Introduction

Current surgical treatment of hair loss mainly relies on transplanting the remaining hair follicles (HFs) to hairless areas [1]. With the advent of tissue engineering and regenerative biology, bioengineering for HF neogenesis is of promising potential for treating such disorders [2]. Before this technology can be applied clinically, a number of properties of the regenerated hairs need to be controlled to restore the natural look of hair patterns, including the number, thickness, spacing, alignment, color, etc. Since hair shows a patterned variation in sizes among body regions and up to thousands of new hairs can be needed in a single patient, a desirable thickness (or the diameter of regenerated hair) and a high regenerative efficiency are vital to a favorable clinical outcome. Prior work has focused more on the preservation of trichogenic ability of cells or the regeneration of HF with a proper architecture [3], [4], [5], [6], [7], [8]. Whether and how the size of regenerated HF can be controlled is yet to be explored. To achieve a reproducible high regenerative efficiency by an engineering approach will also help to bring such technology closer to clinical application.

HF is an ectodermal organ composed of two main parts, including the epithelium of keratinocytes and the mesenchyme of dermal papilla (DP) cells [9]. HF development is regulated through a sophisticated epithelial–mesenchymal interaction [9], [10]. In adult mammals, spontaneous HF neogenesis after injury is rarely observed with only few exceptions, such as deer antler regeneration and a large wound of younger mice [11], [12]. In addition to such spontaneous neogenesis, it was also found that freshly isolated or low-passage cultured DP cells are able to induce HF neogenesis when they are associated with competent epithelial keratinocytes [6], [13], [14]. We and other groups have shown that the HF inductivity of cultured DP cells is better maintained when they are cultured and transplanted as multicellular aggregates, a structure similar to the natural intercellular organization in vivo [15], [16], [17]. Compared with rat or mouse DP cells, human DP cells expanded without Wnt/b-catenin signaling activation are unable to induce HF neogenesis when they are transplanted as dissociated cells [4], [18]. Even freshly isolated human DP cells are not efficient in regenerating new HFs when they were directly transplanted [19]. How to enhance the HF induction efficiency of cultured human DP cells is among the priority issues to be solved in bioengineering for clinical HF regeneration.

To tackle the issue of numbers and to control the spacing of regenerated HFs, we have proposed a 3-step procedure for clinical HF regeneration: expansion of DP cells in vitro, mass production of injectable DP aggregates in a bioreactor, and transplantation of DP aggregates to skin with appropriate spacing and arrangement [15], [16]. We have developed methods for mass production of DP aggregates or HF organ germs through substratum-facilitated self-assembly [15], [16], [20]. Though DP aggregates obtained are able to induce HF neogenesis, there is variation of the size of regenerated HFs [15]. Due to the inhomogeneous sizes of the DP aggregates generated by our previous methods [15], [16], it is unclear whether the variation of the size of regenerated HFs is caused by the size inhomogeneity of transplanted DP aggregates or is a stochastic nature during HF regeneration.

In adult human, the variation of the hair thickness is believed to be controlled by the volume of DP. A larger DP containing more DP cells supports the growth of a larger HF [21]. Comparing the cell numbers of DP and the diameter of hair shafts of rat whiskers, we also found a positive correlation of DP cell numbers with the diameter of hair shafts (Supplementary Fig. S1). Though not tested yet, it is intriguing to speculate that inductive DP aggregates of a larger volume or higher cell numbers may be able to induce neogenesis of larger HFs. This issue has not been systemically investigated, possibly due, at least in part, to the lack of a reliable method for scalable production of size-controllable injectable DP spheroids for reproducible experimental testing. Several methods have been developed to culture dense microtissues, including rotation, two-step rotation and floatation and hanging drop [22], [23], [24]. However, these methods are not suitable for automated scalable production of highly inductive DP spheroids and the ability to tightly control the size and cell number in each spheroid has not been demonstrated. Here, we developed a method that can be automated for mass production of human and rat DP spheroids with wide-range controllable size and cell number. We also examined the effect of the size of DP spheroids on the inductive efficiency and thickness of regenerated hairs.