Microfabrication of a micron-scale microbial-domestication pod for in situ cultivation of marine bacteria

Through the hyphenation of microfabrication, microfluidics and microbiology, we report the development of a μMicrobial-Domestication Pod (μMD Pod). This in situ cultivation device facilitates cell signaling from neighbouring species and interactions with environmental stimuli for marine bacterial growth to overcome current barriers faced by standard laboratory cultivation methods.


Introduction
'The Great Plate Count Anomaly 1 ' is one of the most prevalent mysteries within microbiology 2 which illustrates the impracticality of cultivating the vast majority of microbial species under standard laboratory cultivation techniques. 3,4 Despite the ambiguity of the root cause, a few theories suggest that the uncultivability of microbes stems from the deciency of chemical signaling from neighboring microbes in single cell cultures, a loss of physiological stimuli such as pressure and temperature changes, 5 or prolonged dormancy. 6,7 For over 40 years, natural products have proven to be an unrivalled source for therapeutic agents. 8 As antibiotic resistant strains of microbes increase, so too does the need for the discovery of new microbial species. 9 The development of novel antibiotic therapeutics must not be overlooked as deaths resulting from antimicrobial resistance are currently around 700 000 people globally each year, with a projected increase to 10 million people by 2050 if no further action is taken. 10 Despite the relevancy of microbes, we nd ourselves in an era described as the discovery void 11,12 with rediscovery and replication of known compounds presenting an increasing issue. Whilst developing methods such as molecular networking has shown promising results in defeating the quandary of replication, 13 access to the missing chemical space is still unobtained.
One method developed to combat 'The Great Plate Count Anomaly' is the use of diffusion chambers. The success of the diffusion chambers has been reported to be 300 times greater for the formation of micro-colonies in comparison to traditional plating techniques. 14,15 Diffusion chambers consist of a hollow structure, sealed with lter membranes, which are then situated within the natural environment for in situ incubation. This in situ cultivation allows for cell-to-cell communication and physiological stimuli changes; however, this process requires serial dilution to extinction in order to purify colonies, which signicantly lowers the throughput characteristics of the device.
The isolation chip, more commonly referred to as the iChip, 16 is a simplistic, low-tech device which yields great potential as proved through the isolation of teixobactin from the novel bacterium Eleheria terrae in 2015. 17 The iChip was an evolution of the diffusion chamber which intended to increase the throughput by individually sorting a single bacterium into its own individual well. Although vastly improving throughput, this method may be viewed as a single cell approach in which co-culture could be diminished through the lack of cell-to-cell communication within the individual wells. Our previous work introduced the concept of the Microbial Domestication Pod (MD Pod), 18 a device allowing for single cell isolation while providing the opportunity for cell-to-cell communication. The preparation of the MD Pod was achieved using a microuidic set up to encapsulate a single cell within an agarose microbead. The agarose microbeads were then inserted into the MD Pod which was used for in situ marine sediment cultivation. The MD Pod featured lter membranes which facilitated the entry of nutrients while preventing the entry of other microbes. However, although the original MD Pod was suitable for the in situ cultivation within marine sediment, its size prevented its use within many invertebrates. This study aimed to improve on our current methodology while producing a device capable of in situ cultivation within small marine invertebrates 19 (Fig. 1). Marine invertebrates have long been considered an exciting source of marine natural products, 20,21 oen however, it is not the invertebrate itself which is responsible for the production of the therapeutic agent, but instead, a bacterial species living in symbiosis. 22 Previous studies in the Kerr lab have demonstrated how in situ cultivation within marine invertebrates can be a fruitful source of new compounds with the use of a modied iChip. However, neither the iChip nor the original MD Pod is suitable for in situ cultivation in many smaller species of marine invertebrates and as such, development of a micro-scale device was necessary.
As mentioned, the microbial diversity living within marine invertebrates, octocorals in particular, shows great promise for the discovery of new natural products. However, no previous work has been capable of exploring the in situ cultivation of bacteria within such a conned area. A novel approach is needed to access these resources while facilitating the diffusion of nutrients and chemical signaling. Herein we report the production of the mMD Pod, a device microfabricated through two photon polymerization yielding high-resolution micronsized 23 features negating the use of fragile membranes. This work maintains the benets of a "domestication" period in the bacterium's natural environment simultaneously harnessing the natural symbiotic relationships present through cocultivation within the nursery of the microbeads, 24 re-opening doors for bioprospecting marine invertebrate-associated natural products.

Materials and methods mMD Pod fabrication
The mMicrobial Domestication Pod (mMD Pod) was designed using Solidworks (Dassault Systems, 2020). The design featured two layers of grid-like walls with 60 mm square openings periodically separated in a 60 mm array. The two layers were overlapped with a slight offset to yield 10 mm square through-holes ( Fig. 2a-c). The top of the Pod featured a 500 mm diameter tapered hole for inserting the microbead suspension into the cavity of the Pod. The internal cavity of the 2.850 mm mMD Pod was 1.74 mL which can contain over 1500 microbeads of 200 mm diameter. The base of the Pod had seven octagonal pedestals required for the post-fabrication removal process off the silicon slide. The design was printed using two-photon polymerization technology (Nanoscribe GmbH, Germany). The device was fabricated out of IP-S resin, a biocompatible, non-cytotoxic material used in microuidic applications. The exposure time for a batch of ve mMD Pods was 24 hours using the IP-S resin. Once printed, the Pods were autoclaved before being loaded with the microbead suspension.
To fabricate such a small device, different aspects of the previous devices were taken into consideration. The use of membranes was present in all previous in situ cultivation devices but presented many challenges due to its fragility leading to tears or improper sealing. In this study, two-photon polymerization (Nanoscribe GmbH, Germany) was used to negate the use of membranes and achieve the desired micronsized features. The application of two-photon polymerization enabled the fabrication of a porous device that could facilitate chemical signaling and nutrient diffusion. The mMD Pod featured an open-top for the loading of the cell-encapsulated microbeads. The mMD Pods were then sealed using natural wax (Michaels, USA). Natural wax was selected for sealing of the device as it is a biocompatible material 25 that is suitable for a saltwater environment and inherently quick drying.
mMD Pod seal testing Three mMD Pods were aseptically loaded using a 21-gauge sterilized needle with microencapsulated Sphingomonas phyllosphaerae which were produced using a previously reported microuidic technique 18 (Fig. 2d). This bacterium was selected due to its distinct yellow pigmentation. Each of these mMD Pods were sealed using a natural wax by using forceps to dip the opening of the pod into the wax heated to 70 C and set aside until dry. Once sealed, the mMD Pods were then submerged in Fig. 1 The process of in situ marine bacteria cultivation using the mMD Pod. Marine Broth 2216 (Difco) diluted to 10% strength (dMB) and incubated at room temperature for 7 days without shaking. Three control cultures were performed simultaneously which consisted of S. phyllosphaerae suspended in 10% marine broth.

Incubation of encapsulated S. phyllosphaerae microbeads
Production of microbeads encapsulating S. phyllosphaerae were produced as previously reported. 18 An encapsulation of Streptomyces sp. was also performed using the same method. The microbeads were suspended within a dialysis bag. The incubation within the dialysis bag was performed to ensure that no outside bacterial communities could gain access to the microbeads whilst allowing nutrients to cross the membrane. The microbeads were incubated within the aquarium for a period of 7 days. Aer completion of the incubation period, the microbeads were removed and stained using DMAO in order to assess microbial growth within the microbeads.

Staining for microbead bacterial growth
Aer the completion of the incubation period, the mMD Pod was sacriced onto a glass slide and the debris was washed with 10 mL of sterile water to remove any surface bound microbeads. Cell viability assessments were performed using uorescent Live/Dead bacterial staining (PromoCell, Germany). For the Live/Dead assay, a ratio of 100 : 1 bacterial sample to stain mixture (prepared according to the manufacturer's recommendation) was conducted and viability was assessed by uorescence microscopy using a Revolve4 microscope and a 20Â objective (ECHO, USA). The microbeads were imaged both prior and post incubation to examine colony growth.

Results and discussion mMD Pod fabrication
Aer printing of the mMD Pods, the printability number was found to be 1.05 AE 0.03 (Pr ¼ L 2 /16A) where L is the perimeter of the pore and A is the area of the pore 26,27 (Fig. 2c, e and f). The pedestals on the base of the Pods provided the stability needed during fabrication and transport as the rst shipment of Pods arrived damaged. The new Pods were successfully fabricated and shipped; however, the Pods have an exposure time of 24 hours per ve Pods and the manipulation of these devices proved to be a challenge as the IP-S resin was quite brittle. Future studies will investigate the use of the IP-PDMS and IP-Visio resins which will yield a more exible device. Additionally, the current design featured 10 mm by 10 mm pores; however, we have shown that smaller pore sizes are attainable for this device using two-photon polymerization (Fig. 2c). As concluded from the pressure simulation (details provided in ESI †), the recommended maximum pressure on the mMD Pods to avoid breakage is 350 kPa, which can allow the pods to reach depths of 35 m in saltwater.
mMD Pod seal testing During this experiment, dMB was used as the culture medium to allow for a clear indication of outgrowth if the Pods were not successfully sealed. S. phyllosphaerae was microencapsulated in agarose beads of 250 mm. The Pods were then incubated for 7 days at room temperature before comparison against control cultures. During this trial, bacterial growth could be observed in dMB for the control cultures due to the visible change in broth colour. However, no outgrowth was observed in cultures inoculated with the mMD Pods sealed with sealing wax as there was no visible change in the broth, nor was there any colony formation for the plated broth. Therefore, it was conrmed that the mMD Pod pores are successfully fabricated and the inlet is successfully sealed.

mMD Pod insertion into coral and microbial in situ cultivation
To assess the application of this technique for in situ incubation applications within marine invertebrate hosts, a salt-water aquarium was set up containing the coral Euphyllia glabrescens (AFK Reef Supplies, Canada). In order to assess the mechanical properties of the mMD Pod it was shown that the mMD Pod can be inserted into a coral (Fig. 3b). However, insertion of the mMD Pod into corals with limited coenenchyma whilst possessing large polyps was observed to be challenging. The coenenchyma is the tissue of the coral that connects the colony of polyps. 26 It is within this region that the transfer of nutrients between polyps occurs, and it is hypothesized to contain diverse microbial communities. 26,27 S. phyllosphaerae containing microbeads were incubated for a period of seven days to assess bacterial growth within the beads in a marine environment (as described in the Methods). Following incubation, the microbeads were retrieved, and the agarose microbeads were stained using DMAO. A signicant increase in the number of colonies growing within the microbead post incubation was observed, indicating that the bacteria multiplied during in situ incubation ( Fig. 3c and d). This experiment was also performed using Streptomyces sp. (RKND-216), a marine bacterium, which also experienced a signicant increase in bacterial growth following the incubatory period (Fig. 3e and f).
Upon previous successful trials for the isolation and cultivation of marine bacteria, the need to reduce the size of the Pods was identied to discover new bacteria in different environments. The mMD Pod is currently in its preliminary proof-ofconcept stage; however, future iterations will have to overcome the current fabrication limitations relating to the codependency of the overall device size and smallest pore size. It has been observed that Octocorallia contain diverse communities of bacteria. 28,29 However, this resource has not yet been fully explored due to the physical limitations present. The primary impediment which has prevented in situ cultivation within octocorals is attributable to the small branches of the corals. In the case of Antillogorgia elizabethae 30,31 and Antillogorgia bipinnata, 32 two species of interest, the branch thicknesses range from 1-2 mm. Given the limited space, an altered approach was developed in this proof-of-concept study demonstrating with technological advances two-photon microfabrication presents a viable approach to enable access to the untapped resource of novel natural products within such octocorals.

Conclusions
We have successfully demonstrated how two-photon polymerization printing can be used as a method of fabrication of in situ cultivation devices. Whilst the technology is currently lacking the resolution to create pore sizes of <0.1 mm allowing for the removal of any membranes, this proof on concept displays great potential. The slow growing S. phyllosphaerae was successfully microencapsulated within agarose beads and through the use of live/dead imaging, bacterial growth was observed within the mMD Pod. As technology advances, we believe this is an approach which will provide the opportunity to gain access to currently obtainable microbial dark matter, harnessing synergistic relationships between marine invertebrate associated bacteria and environmental stimuli. This has the potential to be a powerful process for gaining access to new taxa of bacteria, in return, providing access into much needed therapeutic agents. Additionally, the mMD Pod's small size provides great potential creating the possibility of in situ cultivation of 'unculturable' bacteria within the human microbiome. However, future studies must be conducted to expand the capabilities of the device by reducing its overall size and pore size.

Conflicts of interest
The authors declare no conicts of interest.