resilient living materials built by printing bacterial spores · living materials, where cells are...

1

ResilientLivingMaterialsBuiltByPrintingBacterialSpores

LinaM.González1andChristopherA.Voigt1*

1SyntheticBiologyCenter,DepartmentofBiologicalEngineering,MassachusettsInstituteofTechnology,

Cambridge,MA,USA

*CorrespondingandrequestformaterialsshouldbeaddressedtoC.A.V.([email protected])

Keywords: SyntheticBiology,EngineeredLivingMaterial,GeneticCircuit,AdditiveManufacturing,3D

Printing

certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was notthis version posted February 2, 2019. . https://doi.org/10.1101/537571doi: bioRxiv preprint

https://doi.org/10.1101/537571

2

Abstract:

Aroutetoadvancedmultifunctionalmaterialsistoembedthemwithlivingcellsthatcanperformsensing,

chemicalproduction,energyscavenging,andactuation.Achallengeinrealizingthispotentialisthatthe

conditions for keeping cells alive are not conducive tomaterials processing and require a continuous

sourceofwaterandnutrients.Here,wepresenta3Dprinterthatcanmixmaterialandcellstreamsina

novelprintheadandbuild3Dobjects(upto2.5cmby1cmby1cm). Hydrogelsareprintedusing5%

agarose,whichhasalowmeltingtemperature(65oC)consistentwiththermophiliccells,arigidstorage

modulus(G’=6.5x104),exhibitsshearthinning,andcanberapidlyhardeneduponcoolingtopreserve

structural features.SporesofB. subtilis areprintedwithin thematerialandgerminateon itsexterior,

including spontaneously in cracks and new surfaces exposed by tears. By introducing genetically

engineeredbacteria,thematerialscansensechemicals(IPTG,xylose,orvanillicacid).Further,weshow

thatthesporesareresilienttoextremeenvironmentalstresses,includingdesiccation,solvents(ethanol),

highosmolarity(1.5mMNaCl),365nmUVlight,andg-radiation(2.6kGy).Theconstructionof3Dprinted

materialscontainingsporesenablesthelivingfunctionstobeusedforapplicationsthatrequirelong-term

storage,in-fieldfunctionality,orexposuretouncertainenvironmentalstresses.


https://doi.org/10.1101/537571

3

Introduction

Living materials, where cells are embedded within a structural scaffold, pervade the natural world,

examplesbeingwood,bone,andskin.Thecellsprovidedynamicfunctions,includingenergyharvesting,

repair, sensing, and actuation. Natural living materials can survive for years, even millennia, under

fluctuating and stressful conditions. There has been an effort to design so-called engineered living

materials, to artificially combine structural components with cells, including those that have been

genetically engineered1,2. Biomedical applications include the differentiation of human cells to grow

artificialtissueortoserveinlivingmedicaldevices3,4,5,6.Thepotential,however,ismuchgreateracross

many applications, where cells are harnessed for their sensing7, chemical/material production8 self-

healing9, self-powering10, self-cleaning8, and self-gluing11, 12 abilities. Natural microbes have been

harnessedtobuildconcreteandfixcracks9,13,buildpackagingmaterials11,12,andtocontrolbreathability

intextilesbyopeningventsinresponsetosweat14.Engineeredbacteriahavebeendemonstratedtobuild

metalwireswithinelectronics15,16,actasapressuresensorwithinadevice16,anddegradepollutants17.A

keychallengeforlivingmaterialsismaintainingviablecellsforlongtimesoutsideofthelaboratoryunder

extreme,unpredictableconditions.

Livingmaterialsrequiretheorganizationofcellswithinastructuralscaffold.Whileitispossible

toenvisioncellsgrowingtheirownscaffoldovermacroscopiclengthscales,forexamplethegrowthofa

tree,theseprocessesareslowandcurrentlydifficulttocontrolgenetically.Analternativeistouseadditive

manufacturing, also referred to as 3D printing, to organize cellswithin a scaffoldingmaterial at sub-

millimeter resolution over macroscopic length scales. Additive manufacturing has revolutionized

industries,fromarchitecturetoaerospace18-23.Therangeofmaterialsthatcanbeprintedhasexpanded

toincludeceramics24,metals25,26,nylon27,silk28,cellulose29,30,andwoodpolymercomposites31.Thesize

oftheobjectscreatedspanfromnanometerstohouses18.

Incorporating living cells into existing 3D printing platforms is challenging because of toxic

materialsandconditions,suchashightemperaturestoextrudeplasticsorUVlightforcuring32-34.Several

approacheshavebeentakentoaddresstheseissues.Anobjectcanbemadeusingaconventionalprinter

first,afterwhichcellsareintroducedonthesurfaceasabiofilmordiffusedintoahydrogel35,36.Another

approachistosuspendbacteriaevenlythroughoutahydrogelandthenusemultiphotonlithographyto

crosslink barriers to entrap the bacteria in 3D geometries37. These approaches can be used to create

intricatephysicalstructures,butdonotprintthebacteriawithinthematerial.Tothisend,“bioinks”have

beendevelopedthatcombinebacteriaandtheir requirednutrientswith theprecursors that formthe

structuralsupport.Zhaoandco-workersdevelopedabioinkbasedonEscherichiacoliandmicellesthat


https://doi.org/10.1101/537571

4

canbeUV-crosslinkedafterprinting38.Usingthis,theyconstructedhigh-resolution3Dstructures(upto3

cm) anddemonstrated that living cells canperform sensing and computing functions. Another bioink

basedonE.coliwasdevelopedthatutilizestheformationofahydrogelbyalginatewhenitcomesinto

contactwithcalciumchlorideontheprintsurface39.LivingmaterialsbasedonE.coliremainfunctionalfor

severaldaysafterprinting.Studartandco-workersdesignedabioinkbasedonashear-thinninghydrogel

(hyaluronicacid,k-carrageenan,and fumedsilica) thatwasshowntobeable toprintBacillus subtilis,

PseudomonasputidaandAcetobacterxylinum40.AbioinkbasedonB.subtilishasalsobeendeveloped

thatutilizesitsnaturalbiofilmamyloidsfiberstoproduce2Dshapesthatcansurviveat4oCfor5weeks41.

Alloftheseapproachesrequirethatthecellsbemaintainedinahydrogelwithampleaccesstowaterand

replenishednutrientstomaintainafunctionallivingmaterial.

Some bacteria survive under adverse conditions by forming endospores – small spherical

structuresthataredormantandtough.Thecellmembranearchitecture iscomprisedofseveral layers

that include an almost impermeable innermembrane, a germ cell membrane, a thick peptidoglycan

cortex,theoutermembrane,abasementlayer,theinnercoat,theoutercoatandthecrust42.TheDNAis

protected by its tight packing by specialized proteins43-45. Spores are able to survive extreme

environmentalinsults,includinghightemperature,freezing,oxidizingagents,acidandalkalinesolutions,

genotoxic agents, solvents, highpressure,X-rays,g-radiation, andUV-light 46,47. They canalso survive

desiccationbyreplacingwaterwithdipicolinicacidandadoptingawrinkledshapetowithstandosmotic

stress44, 45. Spores can lie dormant indefinitely and purportedly formillions of years48. Spore-forming

bacteria have been used in biotechnology as vaccines, radiation detectors, insecticides, hygrovoltaic

generators,andbio-cements9,10,49-51.

Here, we modify a 3D printer (MakerBot Replicator) that builds objects by extruding plastic

throughahigh-temperaturenozzle(Figure1a).Thenozzleisredesignedtomixtwostreamstoformthe

bioink just prior to printing: one a polymer that is maintained at high temperature and a lower

temperaturemixtureof cellsandmedia. Agaroseexhibits shear thinningandweshowthat it canbe

printedwithvariousthermophilicBacilli species.Usingonlypurifiedspores inthebioink improvesthe

fractionofviablecellsandtheirdistribution in the inprintedstructure.Thesporesremaindistributed

throughout thematerial and provide a constant source of germinated cells at the surface, including

spontaneously in new cracks or tears. There, they can perform their programmed function, such as

respondingtochemicalsdetectedbygenetically-encodedchemicalsensors(shownforIPTG,vanillicacid,

andxylose).Further,thespore-containingmaterialcanbedesiccatedandstoredatroomtemperature

indefinitely.Uponrehydration, theprintedshapereconstitutes, thecellsgerminate,and theyperform


https://doi.org/10.1101/537571

5

their programmed function. When various extreme stresses are applied to the materials, including

ethanol,highosmolarity,UVlight,andradiation,thesporesareabletosurviveandquicklygerminate.

Thisworkdemonstratesaroutebywhichmaterialscanretainthefunctionsprovidedbyembeddedliving

cellslongaftertheyarecreated.

Results

Repurposingofa3DPrinter

TheMakerBot Replicatorwas selected as a simple, inexpensive fused depositionmodeling (FDM) 3D

printer.Theprinterhadtobeextensivelyre-engineeredtoprintlivingcellswithinastructuralmaterial

(Figure1b,SupplementaryFigures1-3).Anovelprinthead(#1inFigure1b)wasdesignedtofacilitatethe

blendingofthestructuralmaterialsandthecellsfromseparatestreamsandtheextrusionofthemixture

througha400µmdiameternozzle.Thestreamsarepropelledusingamodifiedpressurizedtankforthe

material(#2inFigure1b)andaliquidDCpumpforthemediacontainingthecells(#4inFigure1b).The

chamberwhereprintingoccursiscooledtoacceleratematerialhardening.Detailedphotos,schematics,

andelectronicdiagramsforthesedesignsareprovidedintheMethodsandSupplementaryFigures4-11.

TheoriginalMakerBotprintheadhasasingleinputthattakesinasolidplasticfilamentandheats

itto220oC;andalternativecommercialprintheadsweredeemedinappropriateforhandlingcells52,53.We

developedanovelprintheadthatisabletoefficientlymixthetwostreamswhilemaintainingaconstant

temperature (Tchamber) (Figure 1b, Supplementary Figure 4). The extrusion rate from the nozzle is

maintainedclosetotherateatwhichtheMakerBotreelsinfilament.WhentheMakerBotprints,itsmotor

turns in the forward direction. Our printhead detects this motion with an optical sensor and it

subsequentlytriggerstheopeningofsolenoidvalvesthatallowsthecontrolofboththematerialandcell

streams.Thestreamispropelledbypneumaticpressure(0.75-1psi)appliedtothematerialstoragetank

thatiscontrolledbyadigitalelectronicreliefvalve(SupplementaryFigures7-8andTable4).Thisflowis

discontinued when the temperature at the top of the printhead reaches ~63oC and cells have been

introduced (see calibration in Supplementary Figure 9). A second optical sensor detects when the

MakerBot’s motor moves via a trigger and then turns on the cell stream with a 15 ms delay

(SupplementaryFigures10-11).Topromotemixingintheprinthead,thecellinputstreamentersata30°

angleandthecellflowissettobehigh(3.2mL/s).Bubbleformationduringmixing(andpresentinthe

agarose)waseliminatedbyaddinga1mmholeatthetopmostpartoftheprintheadforairtoescape(see

SupplementaryFigure4).


https://doi.org/10.1101/537571

6

TheMakerBotstopsthefilamentfromexitingthenozzlebyturningthemotorbackward;inour

system,this isdetectedbyathirdopticalsensor,which leadstothestoppageofthematerialandcell

streams.Earlydesignsexperiencedproblemswithclogging(notshown).Toavoidthis,atthesametime

thatasolenoidvalvestopstheflowofagaroseneartheprinthead,thereliefvalvereleasespressurein

theagarosetank.

Itiscriticaltomaintainthetemperatureinthelinescarryingagarosetoandattheprintheadso

thatitissufficientlyhightofacilitatetheflowofthematerialstream,yetislowenoughtonotkillthecells.

Forthespecificmaterialandcellcombinationweuse(nextsection)thisbalancewasachievedat75oC

(SupplementaryFigures12-13).Thetemperatureismonitorednearthenozzleoutput(T1),thecenterof

theprinthead(T2)andneartheinputstreams(T3)(Tchamber=<T1,T2>).Tchamberismonitoredusingafeedback

controlsystem(aPIDcontroller,seeSupplementaryFigures14-17).T3isusedtomonitoranddetectwhen

theagaroseisrunninglowandallowagaroseflowviaopeningthesolenoidvalve(s).Thecellsarestored

atT=22oCandexposedtohightemperatureforlessthan20min.Themixtureiscooledimmediatelyafter

exitingthenozzlebymaintainingtheprintchamber(#3inFigure1b)at16oCusingthermoelectriccooling

(peltierelements)(SupplementaryFigures18and19).Thiscoolingsignificantlyimprovesthestructural

detailsintheprintedobject(SupplementaryFigure20).

Toprintanobject,itisdrawnusingSolidWorks,savedinastandardformat(e.g.,asaSTLfile)and

loadedintothesoftwareprovidedbyMakerbot.Thegcodeparametersaremodifiedsothatthenozzle

temperatureis72oC,platformtemperatureis22oC,andtheinfillis100%.Fordualprinting,wemerged

twoSTLfilesusingthemergetoolinReplicatorX(seepictureofthesysteminSupplementaryFigure5).

Mixing is initiated by filling the printhead with a pre-mix of material, running the temperature PID

controllercodetobringthetemperatureintheprintheadto75oC,thenaddingcells(Methods).Afterthis,

theobjectedisprintedfollowingthesameprotocolastheobjectsproducedbytheunmodifiedMakerBot.

MaterialandStrainSelection

Storingtheliquidscontainingthematerialandcellsseparately,asopposedtoasinglebioink,allowsthem

to be independently varied and stored under different optimized conditions. Still, they must be

compatiblewitheachotherastheyaremixedintheprinthead.Thecriticalparameter inourdesign is

temperature,whereamaterialneedstobeselectedthatmeltsatatemperaturethatdoesnotkillthe

cells.Similarly,morethermophilicstrainswillbecompatiblewithawiderrangeofmaterialsthatmeltat

highertemperatures.Aftertestingdifferentmaterialsandbacterialspecies,wefoundthatagaroseandB.

subtilissporesproducedthemostconsistentstructureswiththehighestviabilityofprintedcells.


https://doi.org/10.1101/537571

7

Agarose is a hydrogel, derived from seaweed, consisting of linear polymers of alternating D-

galactoseand3,6-anhydro-L-galactopyranosesubunits.Itisusedcommonlyinbiologyandbeenshown

tosupportembeddedcells54.Agarosewasselectedbecauseitexhibitsshearthinning,istransparent,has

highwatercontent,isstrongafterprinting,andrapidlysolidifieswithouttheneedforchemical/physical

inputsorUV-curing.Whileagarosemeltsat65oC(Figure1c),wefoundthat72oC(Tnozzle) isoptimalfor

obtainingbondingbetweenprintedlayersandavoiddelamination.

Ahigherconcentrationofagaroseleadstoamorerigidstructure,butismoredifficulttoprint.

Wefoundtheidealconcentrationtobe4%(weight/volume).At3%,theagaroseisnotabletoharden

into thedesiredgeometry,evenafter cooling thechamber (SupplementaryFigure21).Above6%,we

started to encounter problemswith line clogging. To balance these effects, thematerial streamwas

chosentobe5%agarosesothatitcanbemixedwiththecellstreamintheprintheadandremain>4%in

the printed object. The storagemodulus (G’) of 5% agarose is» 105, consistentwith the stiffness of

previously published 3Dprinted hyrdrogels17, 55 (Figure 1d and Supplementary Figure 22). The rate at

whichthematerialhardensafterexitingthenozzleiscapturedbytheinitialchangeinG’uponshifting

from70oCto16oC(Figure1d).Inourhands,arate>5000Pa/s(>4%agarose)isimportantforaccurately

printingthedesiredstructure(Figure1dandSupplementaryFigure23).Agaroseisanon-Newtonianfluid

thatexhibitsshearthinning,whichaidsitsextrusionthroughthesmallnozzle(SupplementaryFigure24).

Due to the temperature requirements of themodified printer (Tchamber = 75oC),weneeded to

identify a bacterial species that can survive in the material stream. Mesophiles and thermophiles

representative of different phyla were tested (Anoxybacillus flavithermus, Bacillus amyloquifaciens,

Bacillus licheniformis, Bacillus megaterium, Cupriavidus metallidurans, Escherichia coli, Geobacillus

stearothermophilus, Geobacillus thermoglycosidasius, Gluconacetobacter xylinus, Gluconacetobacter

hansenii, Halobacterium salinarum, Lactobacillus lactis, and Sporosarcina pasteurii) (Supplementary

Figures25-27).Experimentsweredesignedtotestsurvivalofthecellsbyexposingthemto75oCfor20

minutes(Methods)(Figure1e,SupplementaryFigures26-27).Ofthespeciestested,onlyspore-forming

organismswereable tosurvivewith littleday-to-dayvariability.WhenB.subtilisPY79 (aderivativeof

168)56isfirstgrowninspore-inducingmedia(DSM),thisenhancessurvivalandallowsforcellstomore

rapidly recover heat shock (Figure 1e blue lines). Furthermore, we exposed the spores to higher

temperaturefor20min(>75oCandupto100oC)andfoundthatsporescansurviveat100oC,butgrowth

isnotdetecteduntilafter8h(SupplementaryFigure28).

The5%agaroseandB.subtilisPY79inDSMwereselectedasthetwocomponentsofthebioink.

Note that other spore-forming organisms would also work as the cell component of the ink


https://doi.org/10.1101/537571

8

(SupplementaryFigure27).Asimplebarwasprintedconsistingof7layers(2by3by25mm)(Figure1f).

Toaidimaging,anexpressioncassettewheregreenfluorescentprotein(GFP)isconstitutivelyexpressed

wasintroducedintotheB.subtilisgenomeattheamyE locus(Methods).Figure1fshowscomparisons

between the structuresproducedwhenB. subtilis is grownunder conditions favoring vegetative cells

versusspores.Theformershowspunctategrowthdistributedunevenlythroughtheagarosestructure.

Whensporesareprinted,theyareevenlydistributedthroughoutthestructure.

3DPrintingFunctionalLivingMaterials

Figure2showsexamplesofgeometriesthatcanbeprintedbythematerialstreamaloneandtogether

with the cell stream. The minimum feature size is consistent with that obtained by the MakerBot

Replicatorwhenprintingplastics(100µm),whichisdictatedbythe400µmnozzlediameter.Thehydrogel

canbeprintedatascaletheutilizesthemaximumsize inthex-andy-dimensions(10cmby20cm).

Simpleshapescanbeprintedupto1cminthez-dimension,requiringtheprintingof~37layers,each300

µm(atarateof3layers/min).InFigure2i,wecomparetheprintedobjectswiththoseobtainedfroma

high-resolutioncommercially-availableprinter(ProJet6000)andtheshapeandresolutionissimilar.

Therearesomeapplicationswhereitisdesirabletoprintcellsatspecificlocations,ratherthan

throughout the structure. To do this, a second printhead is added intowhich thematerial stream is

injectedviaateeconnector,butforwhichthereisnocellstreaminput(SupplementaryFigure5).TheSTL

filesarepreparedfortheportionscontainingandlackingcellsandthenmerged.Thisgeneratesstructures

withsharpboundariesbetweentheregionscontainingcells(Figure2h).Itisstraightforwardtoexpand

thisapproachtoprintmultiplematerialsandcelltypes.

The spores are distributed evenly throughout the material in which they are printed. They

germinatepreferentiallyatthemedia-exposedsurfaceofthehydrogelduetotheavailabilityofoxygen

andnutrients(Figure3a-candSupplementaryFigure29).Germinationoccursuptoadepthof0.85±0.16

mm,determinedbyimagingcellsconstitutivelyexpressingGFP.Whenthestructureistornandincubated

inLBmedia for10hours, this induces thegerminationofsporesat thesurfaceexposedto themedia

(Figure3c). Thisdemonstratesthattheembeddingofspores inthematerialprovides,undertheright

conditions,asourceofnewgerminatedcellsthatcanperformtheirfunction.

Thegerminatedcellscanfunctionasbiosensors. Todemonstratethis,we implementedthree

genetically-encodedsensorsinthegenomeofB.subtilisthatrespondtosmallmolecules.Thesensors

wereeither taken from the literatureorbuiltdenovo andoptimized forB. subtilis tomaximize their

dynamicrange,suchthattheycanbeimagedinthematerial.AnisopropylB-D-1-thiogalactopyranoside


https://doi.org/10.1101/537571

9

(IPTG)-induciblesystemwasused thathadbeenpreviously reported tohaveadynamic rangeof276-

fold57,58(Figure4aandSupplementaryFigure30).Asugarsensor(xylose)wasmodifiedtofunctioninLB

media by removing the catabolite-responsive element (Figure 4c, Supplementary Figure 31,

SupplementaryTable7)59.Finally,asensorforaplantrootexudate(vanillicacid)wasdesignedbasedon

anoptimizedVanRrepressorandcorrespondingoperator thatwas inserted intothe -10/-35regionof

Pspank(hy)tocreatePspank(V)(Figure4d,SupplementaryTable7)7,60.Thissensorgeneratesa60-folddynamic

range(SupplementaryFigure32).

Materialswereprintedwithsporesofbacteriacontainingthesensors(Figure4).Afterprinting,

thematerialwasincubatedat37oCinmediaeithercontainingorlackingthesmallmolecule.Forallthe

sensors,someinductioncanbeseenat6hourswithfullinductionat12hours.

SporeSurvivalin3DPrintedMaterials

Thesensitivityoflivingcellsandtheircontinuousrequirementfornutrientsandwater,makeitchallenging

to embed themwithinmaterials and function over long times. First,we evaluated the ability for the

printedmaterialstosurvivedesiccation.The5%agarosematerialcontains92%water(Methods).Afterit

isprinted,thewatercanberemovedbydryingandstoredforlongtimes(Figure5a).Whenrehydrated,it

returnstoitsoriginalprintedshape(Figure5b).WhensporesofGFP-expressingB.subtilisareprinted,

theysurvivefor1month(themaximumtested)andgerminatewhenrehydrated.GFPexpressionafter

germinationandgrowthdoesnotgodownforlongerstoragetimes(Figure5c,SupplementaryFigure33).

Given the known longevity of spores61, it is expected that they would be able to regerminate after

indefinitestorageundertheseconditions.

Sporescansurviveundermoreextremeenvironmentalstresses.Alcoholisknowntonothave

sporicidalactivity.Whenprintedobjectscontainingthesporesareexposedto100%ethanolforone

week,thecellssurvive(Figure5dandSupplementaryFigure34).Highosmolarityhasbeenshownto

haveaneffectonsporeviability62.Again,afteroneweek,thesporesremainviable.Notethat,forboth

ofthestresses,someofthisprotectionmaybebeingconferredbytheagarosematerial.

ThematerialscontainingsporeswerethensubjectedtoUVlight,X-rays,andg-radiation(Figure

5dandSupplementaryFigures35-37).Thesporessurvive365nmUVlighttreatment,buttheydonot

survivewhenexposedto254nmlight.Thisisexpectedas254nmUVlightisknowntobeeffectiveat

exterminatingspores63.ThesporesalsosurvivewhenthematerialisexposedtoX-raysfor10min(8.469

R/min).Whenexposedtog-radiation,thesporessurviveafter1hour(2.6kGy)exposure,butnot6hour

(15.7kGy)or12hour (31.5kGy) treatments. Note thatg-radiationwithadoseof25kGy isused for


https://doi.org/10.1101/537571

10

sterilization64.Inaddition,after6hourstheagarosehydrogelweakensandthebarsstarttodeteriorate

duetoshakingduringthegrowthassay(Methods).

Conclusion

Thisworkdemonstratesthatembeddingspores,asopposedtovegetativecells,offersadvantageswhen

producinglivingmaterials.Usingsporessimplifiesthepreparationandstorageofthecellcomponentof

thebioinkandgeneratesmorereliableprintswithouthavingtocarefullycontroltheviabilityorgrowth

phaseofthecellsinthereservoir.Whileweprintthesporeswithinanagarosehydrogel,theyareableto

surviveextremedesiccationandstorageunderambientconditionswithoutacontinuousnutrientsupply.

Whentheygerminate, theyareable toperformtheirprogrammed function,whether itbe tosensea

chemical or turn on a biosynthetic pathway. Further, the spores selectively germinate on the object

surface. Thisallowsthespores inthecenterofthematerialtoprovideacontinuoussourceoffreshly

germinatedcells. Whenthematerial istorn,thesporesgerminateandgrowonthenewsurface.One

couldimagineharnessingthiscapability,alongwiththeabilitytoproducebiopolymers65,tocreateself-

healing materials. While we utilize the natural signals that induce germination, methods have been

developedtogeneticallyengineersporestogerminateunderdefinedconditions,suchasthepresenceof

chemicals,includingneurotransmittersandspecificDNAsequences66.Therecalcitrantnatureofspores

makesthisapproachpotentiallycompatiblewithawiderangeof3Dprintingtechnologybasedonharsher

conditions,suchaslasersinteringofceramics/metals(highT).

Manyapplicationsofsyntheticbiologyrequirethatengineeredcellssurviveandperformtheir

function in fluctuating and stressful environmental conditions67. UV protection will be required for

bacteriathatfunctionintheopenenvironment,forexampleonplantsurfacesinagriculture,68inopen

pondsforbio-production69,orpollutionremediation70.Theabilityofsporestosurvivewithoutwaterand

nutrientsandunderharshconditionscanaid long termstorage.Harnessing the recalcitrantnatureof

spores,stabilizedwithina3Dprintedhydrogelisasteptowardsrealizingtheseapplications.


https://doi.org/10.1101/537571

11

Methods:

Modificationstothe3DPrinter

Wemade extensivemodifications to aMakerBot Replicator 2X, entailing five separate subsystems: a

printhead,anagarosepumpingsubsystem,acellpumpingsubsystem,andaheatingsubsystemanda

coolingsubsystem(SupplementaryFigure1).Thecellsandagarosearepropelledtotheprintheadusing

theagarosepumpingandthecellinputsubsystems.Attheprinthead,thetemperatureisregulatedusing

theheatingsubsystem.Attheotherextreme,thecoolingsubsystemmaintainsthechamberattheproper

temperaturefortheagarosetoretain its filamentaryshapewhen itexits thenozzle.Details regarding

thesemodifications,includingphotos,schematics,partnumbers,filesfor(traditional)3Dprintingofnovel

parts,andArduinoprogrammingdiagramsareprovidedintheSupplementalInformation.

Printhead and optical sensors subsystem. The originalMakerBot printhead has a steppermotor that

extrudestheplasticfilament,drawninasasolidfilamentfromasingleinlet.Ourprinthead’sbarrelhas

twoinlets,oneforthematerialandtheotherforthecells.Italsohasthreethermistorsensorstomeasure

thetemperatureatthetop,middleandbottomoftheprinthead(T1,T2,T3)(SupplementaryFig.4aand

SupplementaryFigure4b).TheaverageofT1andT2isusedtomaintainthetemperatureintheprinthead

at75°Ctoensuremixingoftheagaroseandspores(seecellinputandheatingsubsystemsections).Adual

printingconfigurationisusedbyalternatingbetweentwosolenoidvalvestohavetwodifferentprinthead

printingeitherthematerialaloneorwiththecells(SupplementaryFigure4c).Opticalsensorsareusedto

detectwhichmotormoves and thus,which extruder is active at a given time.Using a pair of optical

interrupters,wecoupledthesteppermotor(fromtheMakerBot)andtheagarosepumpingsystem.An

encoder(10spokes)monitorsthesteppermotor’smovement.Whenamovementchangeisdetectedand

thetemperatureis~63°C(attheverytopoftheprinthead),thesolenoidvalveisopensothatagarose

flowfreelyintotheprinthead.AslotopticalinterrupterisaU-shapedsensorwithaninfraredemitteron

onesideandthecorrespondingdetectorontheoppositeside.Whenanobjectbreakstheinfraredbeam,

thiscanbedetectedthroughabreakinthecircuit.Two8mmwideslotsensorswereplacedintandem

(Digi-KeyElectronicspartno.EE-SX1070).Thesearealsousedtodetectwhenthemotorreversedirection

usingaquadratureencodersignal.Briefly, ifthetwochannelsare90°outofphaseandifpin2seesa

changebeforepin3,thenthemotorismovingclockwise;otherwise,itismovingcounterclockwise(see

code).Asingleopticalinterrupter(Digi-KeyElectronicspartno.EE-SX1042)wasusedtoturnonthecell

streamforduringthetimeittakesfortworevolutionstooccur(seecellinputsubsystemformoredetails).

Ahole(1mmindiameter)wasintroducedatthetopoftheprintheadforexcessairtoescapeafterpriming


https://doi.org/10.1101/537571

12

andwhileprinting(airfromtheagarosecabinetisconstantlypurgedintotheprinthead).Thismodification

relievedtheproblemwithhavingbubblesinsidetheprintedobjectthatdistortedthematerialandledto

non-uniformprints. In SupplementaryFigure6,we showa schematicof theelectronic circuitused to

controlthethreeslotsensors.Eachofthetwoprintheadscontainsthreeopticalsensors.

Agarosepumpsubsystem.Ahermeticallysealeddesiccatorcabinet(FisherScientific,Model#33060)was

repurposedasapneumaticpump(SupplementaryFigure7). Thiscabinethasa¾”thickwallmadeof

polymethylmethacrylate (PMMA).Thevacuumgaugewas replacedwithapressuregauge (McMaster-

Carr,partno.4000K721)witha rangeof0-15psi tomonitorandmaintainaconstantpressure in the

cabinet.Oneorificeinthecabinetwasusedasaninletforhouseairtoenterandtheotherorificewas

used as an outlet for agarose to leave using positive pressure (Supplementary Figure 7a and b). To

maintainagaroseflow,amagnetwire(AppliedMagnets,AWG21)waswoundaroundtheplastictubing

suchthatapassingcurrentreleasesheat(Jouleheating).Tubingconsistedofhigh-temperaturesilicone

rubber.Athickertubing(OD=3/4”,ID=½”,McMaster-Carr,partno.5236K47)wasusedtofittheagarose

outletandastraightreducerbarbedfitting(McMaster-Carr,partno.5463K635)toconnecttoathinner

tubing(OD=5/16”, ID=3/16”,McMasterr-Carr,partno.5236K841)thatfitsthesolenoidvalves.The

lengthofthetubeis27”longtoprovidesomeslackfortheprintheadtomove.Thecurrentinthepower

supplywassetto5.0Aandthevoltageis~12V(BKPrecision1901BSwitchingModePowerSupply).In

addition,thecabinethasanelectricaloutletusedtoplugaheatingtapeformaintainingtheagarosein

thereservoirasaliquid.Thetubingthatcarriesagarosefromthereservoirtotheagaroseoutletonthe

liddoesnothaveanywirewindingsanditstubingissmaller(OD=1/4”,ID=1/8”,McMasterr-Carr,partno.

5236K83).Thereisano-ring(size114)intheinletandoutletofthelidofthecabinettopreventleakage.

Inaddition,acheckvalve(McMaster-Carr,partno.6079T58)preventstheflowofthisviscousliquidback

intothereservoir.Anarrowinreliefonthecheckvalvewassandeddownandthetwolayersofthemagnet

wirewaswoundaroundthisvalve.A1polepush-insignalpowerconnector (McMaster-Carr,partno.

9193T11)wasinstalledformaintenanceassometimesthisvalvegetscloggedandneedstobecleaned.

Epoxywasusedtopermanentlyfixthemagnetwiretothecheckvalves.Thischeckvalveaddsabout1psi

ofresistancetotheline;morecheckvalvesledtoseverecloggingandcleaningdifficulty.Anelectronic

reliefvalvewasinstalledinthecabinettoimmediately(withinafewms)relievethepressurewhenitis

notneededtoactuateflowtotheprinthead.Whenstartingaprint,thelineisinitiallyprimedwithmolten

agarose,whichkeepsthisreliefvalveclosed.Whenthemixingchamber’stemperaturereaches~63°C,

thisindicatesitisfull,andthevalverelievesthepressureinthetank(SupplementaryTable4).Whenthe


https://doi.org/10.1101/537571

13

moltenagaroseispumbedfromthecabinet,thepressureisincreasedandpulsedonandoffsothatit

remainsbetween0.75and1psi.Apairofsolenoidsisplacedbetweenthecheckvalveandtheprinthead

(SupplementaryFigure7aandSupplementaryFigure5)toexertmorecontrolovertheagaroseflowand

to facilitate dual printing (see section on Printhead and slot sensors subsystem for more details).

SupplementaryFigure8showsaschematicoftheelectroniccircuitusedtocontrolthepumpingofthe

agarose.

Cellinputsubsystem. A DC liquid pump (Trossen Robotics, part no. HW-WVALVE) (Supplementary

Figure9a)isusedtopropelthecellstreamintotheprinthead.Theflowismodulatedbyasolenoidvalve

thatiscontrolledbyanArduinomicrocontroller.Toreducethepumpflowrate,8checkvalvesareplaced

intandem,includingoneclosetotheentrypointtotheprinthead(eachcontributing~1psiofresistance)

(SupplementaryFigure9b).Wecalibratedthepumpwiththetimethatthepumpisenergizedoropen

anddeterminetheamountofliquidtoaddforcontinuouspumpingandtokeeptheagarosepercentage

above4%(SupplementaryFigure9c).Aftertheinitialagaroseprimingoftheprinthead,itcontainsabout

5mLofagaroseandweinput600µLofcellstokeeptheagarosepercentageabove4%.Toadd600µLof

cells,weopenedtheDCliquidpump(andsolenoidvalve)for90ms(seeDCliquidpumpcalibrationin

Supplementary Figure 9c). For continuous and subsequent input of the cells, we did a second set of

calibrationstodeterminehowmuchcellstoaddper100µLofagaroseandchose15msforthetimethe

DCliquidpumpandsolenoidvalvesareopened(SupplementaryFigure11).

Heatingsubsystem.Onepowersupply(Newark,partno.78Y9404)iskeptataconstantvoltage/current

andisusedtopowertheheatingofthematerialslinetokeeptheagarosemoltenpriortoreachingthe

printhead.Tomixcellandagarosestreams,thetemperatureintheprintheadneedstobemaintained75

°C(SupplementaryFigure13).Todothis,asecondpowersupply(Newark,partno.34T4663)iscontrolled

remotelyandweimplementaPIDcontrollertomaintainthisconstanttemperatureintheprinthead.The

PIDalgorithm(SupplementaryFigure14a)comparesthesetpointandmeasuredtemperaturesandused

theirdifference(theerror)tocontrolthecurrentfromthepowersupplybasedontheequation

𝑜𝑢𝑡𝑝𝑢𝑡 = 𝐾'𝑒 𝑡 + 𝐾+ 𝑒 𝑡 𝑑𝑡 + 𝐾--(/(0)-0

(1)

wheretheoutputisthesignalcontrollingthecurrent,tisthetime,Kp,KiandKdarethecoefficientsfor

theproportional, the integraland thederivative, respectively. It is implementedusinganArduinoPID

library and the tuningparameters are adjustedusing an autotune function (SetTuning). The feedback

controlisimplementedusingtheremote-controlledpowersupplywithan8pinplugadapter(onlypin1


https://doi.org/10.1101/537571

14

(5V),pin2(input)andpin3(ground)wereused),magnetwire,thermistorsandanArduinomicrocontroller

(SupplementaryFigure14b).SupplementaryFigure15showsaschematicofthecircuitwiring.Weverified

thatthesystemandcodeworkatroomtemperature(SupplementaryFigure16)andwhilecoolingthe

chamber(SupplementaryFigure17).Thetemperatureofthenozzleisindependentoftheprintheadand

wecontrolleditthroughthegcodeintheoriginalMakerBot.Wescreenedforthebesttemperatureto

keep the nozzle and we found it to be 72°C (Tnozzle + Tchamber + Tambient) which leads to an average

temperatureof~54°C.

Cooling subsystem. The rapid gelation of the agarose/cell mixture after leaving the printhead nozzle

requiredthecoolingoftheprintchamberto14-17°Cbyblowingcoldairintoit.AStyrofoamboxwithan

aluminumpanwasplacedbehindtheprinter(SupplementaryFigure18).Twometalbarsthatspanacross

thepanwereattachedtopeltierplates(SupplementaryFigure19).Afanblowsairintothecoolingsystem

andtwoaluminumelbowpipesmovethecoldairintotheprintingchamber.Metalfinswereattachedto

thebarstoincreasethesurfaceareaandfacilitateheattransferasairpassesthroughthesefins.Priorto

usingtheprinter,1Lofwater isaddedtothealuminumpan, it iscooledforonehour,andthen ice is

addedtothepan.

Preparationandstorageofagarose.TheSeaPlaqueagarose(Lonza,catno.50100)powderisdissolvedin

300mLofwater(tomakea5%w/v),autoclaveandstoredatroomtemperature.Rightbeforeaddingit

to the material reservoirs (a 600 mL beaker with a heating tape wrapped around it) the agarose is

microwavedfor3.5minoruntilitcompletelymelts.Wecoverthebeakerwithapieceofaluminumto

preventexcessiveevaporationinsidethedesiccatorcabinet.

Embeddingandscaffoldingmaterialscharacterization. Usingarheometer(AR2000,TAInstruments),we

characterized the gelation time, the melting temperature, and the shear thinning properties of the

agarose.Allofthemeasurementsweretakenusingapeltierplate(withelementsforrapidcoolingand

heating)(itcanrapidlycoolandheat).Toobtainadequaterheometermeasurementsofthehydrogels,a

cross-hatchedgeometry(543337.001)wasused.Thisprovidesaroughsurfacetogriponthehydrogels.

Toavoidslippingofthehydrogelonthemetalsurfaceofthepeltierplateandprovidesometexture,a

self-adhesivesandpaper(320grit,cat.McMaster-Carr,no.4647A13)wasused(SupplementaryFigure

22a).Topreventdryingofthehydrogels,weusedanevaporationsolventtrap(SupplementaryFigure22;

inset).The%strainwasset to0.1and the frequencyat1Hz forallmeasurements.Thesampleswere


https://doi.org/10.1101/537571

15

preparedasintheprevioussection,butinasmallervolume(5mL).Torunthesamplesontherheometer,

about1mLispipettedontotheself-adhesivesandpaper,thegeometryisloweredtoabout1050µmfrom

thesurfaceandletsolidify.Theexcessmaterialisthenremovedusingaspatulasothatcross-sectional

areamatchesthatofthegeometry(490mm2)andthisgeometryisthenlowerto1000µm(orspecified

gapinthefile).

MolecularBiologyandGeneticEngineering

Strainsandcultureconditions.ForacompletelistofstrainsusedinthisstudyseeSupplementaryTable5.

Chemically-competentEscherichiacoliDH10Bcellswereusedforallroutinecloning(NewEnglandBiolabs,

Ipwich,MA).TheE.coliMG1655(inFigure1e)andthevariousBacillusstrainwereobtainedfromthelab

stockoriginallyobtainedfromeitherATCCorBacillusGeneticStockCenter(BGSC).AllE.coliandBacillus

strains were grown at 37oC in LB-Miller media (BD, cat no. 244620) unless indicated. Anoxybacillus

flavithermusstrainy.d.(DSM2641)cellsweregrownat55oCinNutrientBroth(5g/Lofpeptone(BD,cat

no.211677),3g/Lofbeefextract(BD,catno.212303),pHadjustedto7.0)andAnoxybacillusflavithermus

strainWK1(DSMZ,Germany)cellsweregrownat55oCinCasoBroth(15g/Lofpeptonefromcasein(BD,

catno.211921),5g/Lofpeptonefromsoybean(Sigma-Aldrich70178-100G),5g/mLofNaCl).Geobacillus

thermoglucosidasiusM5EXG(ATCCBAA-1069)weregrownat55oC inTrypticSoyBroth(40g/L,BDcat.

211825).Lactococcuslactissubsp.cremoris(Mobitech,MG1363),Lactococcuslactissubsp.lactis(ATCC

11454)weregrowninBrainHeartInfusion(BHI)Broth(BD,237500).Gluconacetobacterxylinus(ATCC

700178)andGluconacetobacterhasenii(ATCC23769)weregrowninHestrinandShramm(HS)medium

at30°C.HScontains2%(w/v)ofglucose(FisherScientific,C6H1206)0.5%(w/v)ofyeastextract(BD,cat

no.210929),0.5%(w/v)peptone(BD,catno.211677),0.27%(w/v)ofsodiumphosphasediabasic(Fisher

Scientific,S375-500),0.15%(w/v)ofcitricacid(FisherScientific,A940-500)and0.1%(w/v)ofcellulase

(Sigma-Aldrich,catno.C2730-50ML).Theglucoseandthecellulasewerefiltersterilizedusinga0.2µm

Acodisc syringe filter (Pall, cat no. PN 4612). Cupriavidus metallidurans CH34 (ATCC 43123) and a

domesticatedstrainderivedfromSporosarcinapasteurii(ATCC11859)weregrowinNutrientBroth(BD

cat.234000)at30°C.TogrowtheS.pasteurii,thepHoftheNutrientBrothwasadjustedto8.5before

autoclaving.HalobacteriumsalinarumNRC-1(ATCC700922)weregrownin2185HalobacteriumNRC-1

mediumat30°C.The2185HalobacteriumNRC-1mediumismadewithaBasalmediumandtracemetals,

asdescribed in theATCCprotocol.MinimalS750mediumcontains the following ingredients ina total

volumeof100mL:10mLof10XS750salts,1mLof100Xmetals,2mLof50%(w/v)arabinose(Sigma-

Aldrich,catno.A3256-100G),1mL10%glutamate(FisherScientific,catno.A125-100).The10XS750salts


https://doi.org/10.1101/537571

16

consistof(filtersterilized):0.5MMOPS(freeacid)(FisherScientific,catno.ICN19483725),100mMof

ammonium sulfate (NH4)2SO4 (Millipore Sigma, cat no. AX1385-1), 50 mM of potassium phosphate

monobasic (KH2PO4) (Sigma Aldrich, cat no. 795488-500G), 8g of potassium hydroxide (KOH) (Sigma-

Aldrich, cat no. 221473-500G) (finely adjusted to pH 7 with 1M KOH). The 100X metals consist the

followingmetals(filtersterilized):0.2Mofmagnesiumchloride(MgCl2)(Sigma-Aldrich,catno.M8266-

100g),70mMofcalciumchloride(CaCl2)(Sigma-Aldrich,catno.C1016-500G),5mMofmanganese(II)

chloride(MnCl2)(Sigma-Aldrich,catno.244589),0.1mMofzincchloride(ZnCl2)(Sigma-Aldrich,catno.

Z0173),100mgmL-1ofthiamine-HCL(Sigma-Aldrich,catno.T4625-25G),2mMofHCL(MilliporeSigma,

catno.HX0603-3),and0.5mMofiron(II)chloride(FeCl3)(Sigma-Aldrich,catno.701122-1G).ForE.coli,

thefollowingantibioticconcentrationsareused(bothplatesandinculture):100μg/mlampicillin(Ap,

GoldBio;CAS#69-52-3);50μg/mlkanamycin(Kn,GoldBio;CAS#25389-94-0),100μg/mlspectinomycin

(Sp, GoldBio; CAS#22189-32-8) and 35 μg/ml chloramphenicol (Cm, AlfaAesar; #25-75-7). For B.

subtilis,thefollowingconcentrationwereused:5μg/mlCm,100μg/mlSp,5μg/mlkn.

Plasmidsandstrainconstruction.AllplasmidswereconstructedusingTypeIISassembly.DNAsequences

wereinspectedandmodifiedtoeliminatetherecognitionsitesfortherestrictionenzymes(BsaIorBbsI

orBsmbI)asnecessary.AbsaIsitewasremovedfromtheoriginalgfpmut2anditwasnamedgfpmut2x.

DNAwasamplifiedusingthepolymeraseKapaHiFiDNApolymerase(KapaBiosystems,catno.KK2602)

andgelpurifiedusingSeakemGTGagarose (fornucleicacidrecovery) (Lonza,catno.50070),agarose

dissolvingbuffer(ABD)(ZymoResearch,catno.D4001-1-50),Zymo-SpinI(ZymoResearch,catno.C1003-

50)andcollectiontubes(ZymoResearch,catno.C1001-50).Allgenetically-modifiedB.subtilisstrainsare

based on Bacillus subtilis PY79 (obtained from the lab stock). All DNA designed for B. subtilis were

integrated into the genome at the amyE site. Genomic integrationwas verified by polymerase chain

reaction(PCR).

PreparationofB.subtilisspores.ToinduceB.subtilissporulation,weusedtheDifcoSporulationMedium

(DSM)containing8g/LofDifcoNutrientBroth(BD,catno.234000),1g/Lofpotassiumchloride(Sigma-

Aldrich,catno.P5405-500g),0.25g/Lofmagnesiumsulfateheptahydrate(MgSO4×7H20)(Sigma-Aldrich,

catno.230391-500).ThefollowingsolutionswereaddedtoDSMafterautoclaving:1Mcalciumnitrate

hydrate (Ca(NO3)2) (Sigma-Aldrich, cat no. 202967-10G), 0.01MMnCl2(Sigma-Aldrich, cat no. 244589-

500G)and1mMiron(II)sulfate(FeSO4)(Sigma-Aldrich,215422-5G)(to1L,1mLofeach).B.subtilisstrains

werestreakedonLBplates(withcorrespondingantibiotics)andsinglecoloniesweregrownintwoculture


https://doi.org/10.1101/537571

17

tubes(in3mLofDSM)for3hpriortotransferringthisinoculumto500mLin2800-LErlenmeyerflask.

Thesecellsweregrownfor2days,shakingat250rpm,at37°C,andinaNewBrunswickScientificInnova

44 incubator.TheOD600wasmeasuredtobe~1.Weverifiedthepresenceofspores inthecultureby

microscopy.Thecellswerestoreat4°Cpriortouse intheprinter.Thismediawastransferredtoa1L

Nalgenecentrifugebottles(FisherScientific,05562-25)forloadingintotheprinter.Thelidofthesebottles

weremodifiedbydrillingahole(0.25”indiameter)tofitatubing(McMaster-Carr,catno5236K83).The

otherendofthetubingwastightlyfitintotheinletoftheDCliquidpump.

Heatshockandoutgrowthexperiments.Athermocycler(Bio-RadC1000Touch)wasusetoheatshock

thecellsfor20minat75°C.Cellswerestreakedinplatesthensinglecolonieswerepickedandgrownin

theirstrain-specificmediaandtemperaturepriortodilution.Thecellswereheatshockedin50µL(OD600

=0.04)andtransferredtoplates(Nunc96-WellOptical-BottomPlate)containing150µLoftheirrespective

medium(finalofOD600=0.01).Thecellsweregrowninamulti-modemicroplatereader(Biotek,Synergy

H1)for20hourswithcontinuousorbitalshakingandattherespectivestraintemperature.

MeasurementofGFPin3DPrintedParts.Theprintedobjectsweregrownin5mLofLBandincubated

withshaking(250rpm)at37°C.Thepartswerewashedwith12mLof1Xphosphate-buffered

saline(PBS)toreducetheautofluorescenceoftheLBmedia.TheexcessPBSwasgentlyremoved

usingKimwipes(Kimberly-Clark™Professional34120).TheobjectswerethenimagedwithaChemiDoc

MPImagingSystem(BioRad,Hercules,CA)usingtheAlexa488filteranda0.5secofexposuretime.

TheplasmidpLG166wasusedtomakethe+GFPB.subtilisstrainLMG09(SupplementaryFigure33a).

Thegfpmut2xgeneisdrivenconstitutivelywiththepromoterPpenandaspectinomycincassettewas

usedforselection.TheplasmidpLG173wasusedtomakethe-GFPB.subtilisstrainLMG16

(SupplementaryFigure33b).ToquantifyandtodotheimageanalysisofthedatashowninFigure4and

5,weloadedtherawimagesintoaroutinemacroforImageJ(availableuponrequest).Thereported

meanfluorescencevaluesarethe+Gfpcells.

Dryingandrehydrationexperiments.Thebareagarosematerialswereprintedanddriedovernightat

room temperature. A green dye was used to color the agarose (Figure 5a). The dried object was

hydratedwithdeionize autoclavedwater. Thewidth and theheightweremeasuredusing a digital

caliper(Global,item#WG534250).Forthelongevityexperiments,wedriedthematerialsovernightin

petri dishes, then wrapped them with 2” parafilm (VWR, cat no. 52858) and stored at room


https://doi.org/10.1101/537571

18

temperaturefor1day,1weekand1month.Forthehydrationpartoftheexperiment,wecompletely

submergedthe3Dprintedpartsin1Xphosphate-bufferedsaline(PBS)for2hourspriortotransferring

the3Dobjectstotheroutineculturingconditionsasdenotedintheprevioussection(Measurementof

GFPin3DPrintedParts).Tocalculatethewatercontent(72%),theabsolutevalueofthedifferenceinthe

wetweightandthedryweight(afterdryingfor24hoursatroomtemperature)ofthehydrogelisdivided

bythewetweightmultipliedby100%.

Inducible systems characterization in culture. Chemical inducers were used in the following

concentrationsunlessindicated:1mMofisopropyl-β-D-thiogalactoside(IPTG)(GoldBio;CAS67-93-1),

1mM of vanillic acid (Sigma CAT 94770) and 1% (w/v) xylose. Prior to using the instrument, single

colonieswereinoculatedin2mLofLB(withtherespectiveantibiotic)andgrownat37°Cwith250rpm

shakingforapproximately3h.ThecellswerebackdilutedtoOD600of0.001andgrownfor1hour(oruntil

earlylogphase)inplates(CostarAssay96-wellplateclearroundbottom)usinganElmishakerat1000

rpmat37°C.After this timeelapsed the corresponding inducerswereaddedand then the cellswere

allowedtogrowfor1.5h.

Inducible systems characterization in material. The same culture conditions as described in the

MeasurementofGFPin3DPrintedPartssectionwasused.Theonlydifferenceisthattheinducerswere

addedatt=0hr.Theconcentrationofinducersaddedforeachsensorwasthesameasinculture(see

Inducible systemscharacterization in culture). The strainsused in this setofexperimentswere theB.

subtilis LMG04 (IPTG inducible system),B. subtilis LMG117 (the vanillic acid inducible system), andB.

subtilisLMG125(thexyloseinduciblesystem).

Sporesinthecoreofthematerials.Blockswithdimensionof6x6x25mmwereprintedusingthespore

oftheB.subtilisstrainLMG09(gfpconstitutivelyactive)andtheB.subtilisstrainLMG16.Thecellswere

grownasindicatedinthesectionMeasurementofGFPin3DPrintedParts,except,thatthemediawas

replacedevery3handthecellswereincubatedforatotalof10hours.Theblockswerecutinhalfusing

disposablescalpel(ElectronMicroscopyScience,catno.72042-21)priortoincubation.Thecontrolgroup

weretheintactblocks(nocutinhalf).Atthe10htheblockswerewashedwith1XPBS.A1mmthinslide

wascutfromeitherthecenterortheedgeoftheblocks.Fortheexperimentalgroup,a1mmthinslide

fromtheedgewascutandforthecontrolgroupa1mmslidewascutfromthecenteroftheblock.For

theimageanalysis,weusedthestraightlineandplotprofiletoolsinImageJ(usingcommandK).Thegray


https://doi.org/10.1101/537571

19

scalevalueobtainedisnormalizedbydividingthevaluesby255(denotedas‘Intensity’inFigure3).

Flowcytometry.CytometrywasperformedwithaLSRIIFortessa(BDBiosciences,SanJose,CA).20µLof

B.subtiliscellsweretransferredto180µL1XPBSwith34µg/mLofchloramphenicol.Foreachsample,at

least20,000eventswerecollected.ThedatawereanalyzedusingFlowJo(TreeStar, Inc.,Ashland,OR).

Thepopulationsweregatedbasedonforwardandsidescatterandthegeometricmeanwasrecorded.

Whitecells(cellsnotproducingGfp,butcontainingaSpcassette)wererunduringeachexperimentto

subtractthecellularautofluorescence.

EthanolandNaClexposureexperiments.Theprintedbarscontainingsporeswereexposedthemto100%

ethanoland1.5MNaCl.Weadded25mL(oruntilsubmerged)ofeithertopetridishescontainingthe

printedpartsandcoveredthemforthedurationindicated(SupplementaryFigure34).Afterwhich,we

washedtheobjectswith12mLof1XPBSpriortoculturingthecellsasdescribedinMeasurementofGFP

in3DPrintedParts.

ExperimentswiththeUltraviolet(UV)light,X-rayandtheGammaIrradiators(SupplementaryFigure34-

37). An Ultraviolet (UV) lamp (Model: ENF-240C; Spectroline™,WestBury, NY) with dual wavelength

(365nmand254nm)wasusedfortheUVexperiments.The lampwasheldwithastand(Spectroline™

SE140)at60mmfromtheplatewiththeparts.TheX-raymachineusedtodothisexperiment isare-

purposedhome-builtsystem(MITRPPX-raymachine).Theoperationalrangeisfrom75-250kVandthe

irradiateddiameteris~6inches.Thefilterusedtoperformtheseexperimentswasa0.2mmCufilterand

wemeasuredthedoseratetobe8.282R/minusinganionbeam(Cat.10X6,RadCal,Monrovia,California).

To do the g-radiation experiments, two gamma irradiators were used: GC-40E (dose rate was 93.16

rad/min)andGC-220E(doseratewas4375rad/mininthemonthofDecember2018).Thetimecourse

was initiated immediately following the exposure period and then the experiments described in

MeasurementofGFPin3DPrintedPartssectionwereperformed(excepttheLBmediawasreplacedand

imagetakenevery3h).


https://doi.org/10.1101/537571

20

Acknowledgments. LMG and CAVwere supported by the Vannevar Bush Faculty Fellowship program

sponsoredbytheBasicResearchOfficeoftheAssistantSecretaryofDefenseforResearchandEngineering

andfundedbytheOfficeofNavalResearchthroughgrantnumberN00014-16-1-2509.Thisresearchwas

alsofundedbytheInstituteforCollaborativeBiotechnologies(ICB)throughcontractnumberW911NF-

09-0001withtheU.S.ArmyResearchOffice.

Printfiles.ThenameoftheSTLfilesoftheprintdesignsareprovidedintheSupplementalFiles.

Dataavailability.Thedatathatsupportsthefindingsofthisstudyareavailablefromthecorresponding

authoruponrequest.

References:

1. Nguyen,P.Q.,Courchesne,N.-M.D.,Duraj-Thatte,A.,Praveschotinunt,P.&Joshi,N.S.EngineeredLivingMaterials:ProspectsandChallengesforUsingBiologicalSystemstoDirecttheAssemblyofSmartMaterials.AdvancedMaterials30,1704847(2018).

2. Chen,A.Y.,Zhong,C.&Lu,T.K.EngineeringLivingFunctionalMaterials.ACSSyntheticBiology4,8-11(2015).

3. Kolesky,D.B.,Homan,K.A.,Skylar-Scott,M.A.&Lewis,J.A.Three-dimensionalbioprintingofthickvascularizedtissues.ProceedingsoftheNationalAcademyofSciences113,3179(2016).

4. Markstedt,K.etal.3DBioprintingHumanChondrocyteswithNanocellulose–AlginateBioinkforCartilageTissueEngineeringApplications.Biomacromolecules16,1489-1496(2015).

5. Yanagawa,F.,Sugiura,S.&Kanamori,T.Hydrogelmicrofabricationtechnologytowardthreedimensionaltissueengineering.RegenerativeTherapy3,45-57(2016).

6. Mao,A.S.&Mooney,D.J.Regenerativemedicine:Currenttherapiesandfuturedirections.ProceedingsoftheNationalAcademyofSciencesoftheUnitedStatesofAmerica112,14452-14459(2015).

7. Meyer,A.J.,Segall-Shapiro,T.H.,Glassey,E.,Zhang,J.&Voigt,C.A.Escherichiacoli“Marionette”strainswith12highlyoptimizedsmall-moleculesensors.NatureChemicalBiology(2018).

8. Gerber,L.C.,Koehler,F.M.,Grass,R.N.&Stark,W.J.Incorporatingmicroorganismsintopolymerlayersprovidesbioinspiredfunctionallivingmaterials.ProceedingsoftheNationalAcademyofSciences109,90-94(2012).

9. Jonkers,H.M.,Thijssen,A.,Muyzer,G.,Copuroglu,O.&Schlangen,E.Applicationofbacteriaasself-healingagentforthedevelopmentofsustainableconcrete.EcologicalEngineering36,230-235(2010).

10. Chen,X.,Mahadevan,L.,Driks,A.&Sahin,O.Bacillussporesasbuildingblocksforstimuli-responsivematerialsandnanogenerators.NatNano9,137-141(2014).

11. Zeller,P.&Zocher,D.Ecovative’sBreakthroughBiomaterials.FungiMagazine5,51-56(2012).12. Haneef,M.etal.AdvancedMaterialsFromFungalMycelium:FabricationandTuningofPhysical

Properties.ScientificReports7,41292(2017).


https://doi.org/10.1101/537571

21

13. Tarczewski,R.Formationofsustainableinfrastructureusingmicrobialmethodsandhumanizationofman-madeenvironment.ProcediaManufacturing3,1704-1711(2015).

14. Wang,W.etal.Harnessingthehygroscopicandbiofluorescentbehaviorsofgeneticallytractablemicrobialcellstodesignbiohybridwearables.ScienceAdvances3,e1601984(2017).

15. Chen,A.Y.etal.Synthesisandpatterningoftunablemultiscalematerialswithengineeredcells.NatMater13,515-523(2014).

16. Cao,Y.etal.Programmableassemblyofpressuresensorsusingpattern-formingbacteria.NatureBiotechnology35,1087(2017).

17. Schaffner,M.,Rühs,P.A.,Coulter,F.,Kilcher,S.&Studart,A.R.3Dprintingofbacteriaintofunctionalcomplexmaterials.ScienceAdvances3,eaao6804(2017).

18. Gosselin,C.etal.Large-scale3Dprintingofultra-highperformanceconcrete–anewprocessingrouteforarchitectsandbuilders.Materials&Design100,102-109(2016).

19. Conner,B.P.etal.Makingsenseof3-Dprinting:Creatingamapofadditivemanufacturingproductsandservices.AdditiveManufacturing1-4,64-76(2014).

20. Joshi,S.C.&Sheikh,A.A.3Dprintinginaerospaceanditslong-termsustainability.VirtualandPhysicalPrototyping10,175-185(2015).

21. Ventola,C.L.MedicalApplicationsfor3DPrinting:CurrentandProjectedUses.P&T:apeer-reviewedjournalforformularymanagement39,704-711(2014).

22. Murphy,S.V.&Atala,A.3Dbioprintingoftissuesandorgans.NatureBiotechnology32,773(2014).

23. Dawood,A.,Marti,B.M.,Sauret-Jackson,V.&Darwood,A.3Dprintingindentistry.Bdj219,521(2015).

24. Feilden,E.etal.3DPrintingBioinspiredCeramicComposites.ScientificReports7,13759(2017).25. Wang,Y.M.etal.Additivelymanufacturedhierarchicalstainlesssteelswithhighstrengthand

ductility.NatureMaterials17,63(2017).26. Martin,J.H.etal.3Dprintingofhigh-strengthaluminiumalloys.Nature549,365(2017).27. Mostafa,K.G.,Montemagno,C.&Qureshi,A.J.StrengthtocostratioanalysisofFDMNylon12

3DPrintedParts.ProcediaManufacturing26,753-762(2018).28. Dickerson,M.B.etal.3DPrintingofRegeneratedSilkFibroinandAntibody-Containing

MicrostructuresviaMultiphotonLithography.ACSBiomaterialsScience&Engineering3,2064-2075(2017).

29. SydneyGladman,A.,Matsumoto,E.A.,Nuzzo,R.G.,Mahadevan,L.&Lewis,J.A.Biomimetic4Dprinting.NatureMaterials15,413(2016).

30. Pattinson,S.W.&Hart,A.J.AdditiveManufacturingofCellulosicMaterialswithRobustMechanicsandAntimicrobialFunctionality.AdvancedMaterialsTechnologies2,1600084(2017).

31. LeDuigou,A.,Castro,M.,Bevan,R.&Martin,N.3Dprintingofwoodfibrebiocomposites:Frommechanicaltoactuationfunctionality.Materials&Design96,106-114(2016).

32. Cazón,A.,Morer,P.&Matey,L.PolyJettechnologyforproductprototyping:Tensilestrengthandsurfaceroughnessproperties.ProceedingsoftheInstitutionofMechanicalEngineers,PartB:JournalofEngineeringManufacture228,1664-1675(2014).

33. Jacobs,P.F.Rapidprototyping&manufacturing:fundamentalsofstereolithography.(SocietyofManufacturingEngineers,1992).

34. Tumbleston,J.R.etal.Continuousliquidinterfaceproductionof3Dobjects.Science347,1349(2015).

35. Keating,S.J.etal.3DPrintedMultimaterialMicrofluidicValve.PloSone11,e0160624-e0160624(2016).

36. Schmieden,D.T.etal.PrintingofPatterned,EngineeredE.coliBiofilmswithaLow-Cost3DPrinter.ACSSyntheticBiology7,1328-1337(2018).


https://doi.org/10.1101/537571

22

37. Connell,J.L.,Ritschdorff,E.T.,Whiteley,M.&Shear,J.B.3Dprintingofmicroscopicbacterialcommunities.ProceedingsoftheNationalAcademyofSciences110,18380(2013).

38. Liu,X.etal.3DPrintingofLivingResponsiveMaterialsandDevices.AdvancedMaterials30,1704821(2017).

39. Lehner,B.A.E.,Schmieden,D.T.&Meyer,A.S.AStraightforwardApproachfor3DBacterialPrinting.ACSSyntheticBiology6,1124-1130(2017).

40. Schaffner,M.,Rühs,P.A.,Coulter,F.,Kilcher,S.&Studart,A.R.3Dprintingofbacteriaintofunctionalcomplexmaterials.ScienceAdvances3(2017).

41. Huang,J.etal.ProgrammableandprintableBacillussubtilisbiofilmsasengineeredlivingmaterials.NatureChemicalBiology15,34-41(2019).

42. McKenney,P.T.,Driks,A.&Eichenberger,P.TheBacillussubtilisendospore:assemblyandfunctionsofthemultilayeredcoat.NatureReviewsMicrobiology11,33(2012).

43. Setlow,B.&Setlow,P.Bindingofsmall,acid-solublesporeproteinstoDNAplaysasignificantroleintheresistanceofBacillussubtilissporestohydrogenperoxide.AppliedandEnvironmentalMicrobiology59,3418(1993).

44. Setlow,B.,Atluri,S.,Kitchel,R.,Koziol-Dube,K.&Setlow,P.RoleofdipicolinicacidinresistanceandstabilityofsporesofBacillussubtiliswithorwithoutDNA-protectivealpha/beta-typesmallacid-solubleproteins.Journalofbacteriology188,3740-3747(2006).

45. Sahin,O.,Yong,E.H.,Driks,A.&Mahadevan,L.PhysicalbasisfortheadaptiveflexibilityofBacillussporecoats.JournaloftheRoyalSociety,Interface9,3156-3160(2012).

46. Setlow,P.SporesofBacillussubtilis:theirresistancetoandkillingbyradiation,heatandchemicals.JournalofAppliedMicrobiology101,514-525(2006).

47. Moeller,R.etal.RolesoftheMajor,Small,Acid-SolubleSporeProteinsandSpore-SpecificandUniversalDNARepairMechanismsinResistanceofBacillussubtilisSporestoIonizingRadiationfromXRaysandHigh-EnergyCharged-ParticleBombardment.JournalofBacteriology190,1134(2008).

48. Cano,R.J.&Borucki,M.K.Revivalandidentificationofbacterialsporesin25-to40-million-year-oldDominicanamber.Science268,1060(1995).

49. Cutting,S.M.,Hong,H.A.,Baccigalupi,L.&Ricca,E.OralVaccineDeliverybyRecombinantSporeProbiotics.InternationalReviewsofImmunology28,487-505(2009).

50. Yung,P.T.&Ponce,A.Faststerilityassessmentbygerminable-endosporebiodosimetry.Appliedandenvironmentalmicrobiology74,7669-7674(2008).

51. Adler,J.&Templeton,B.TheEffectofEnvironmentalConditionsontheMotilityofEscherichiacoli.JournalofGeneralMicrobiology46,175-184(1967).

52. N.Turner,B.,Strong,R.&A.Gold,S.Areviewofmeltextrusionadditivemanufacturingprocesses:I.Processdesignandmodeling.RapidPrototypingJournal20,192-204(2014).

53. Hongxi,Y.,Ming,Q.,Haiyan,Z.&YeChan,L.3DPrintingandBuildings:ATechnologyReviewandFutureOutlook.Technology|Architecture+Design2,94-111(2018).

54. Levskaya,A.,Chevalier,A.A.,Tabor,J.J.,Simpson,Z.B.&Lavery,L.A.Syntheticbiology:engineeringEscherichiacolitoseelight.Nature438(2005).

55. Compton,B.G.&Lewis,J.A.3D-PrintingofLightweightCellularComposites.AdvancedMaterials26,5930-5935(2014).

56. McLoon,A.L.,Guttenplan,S.B.,Kearns,D.B.,Kolter,R.&Losick,R.Tracingthedomesticationofabiofilm-formingbacterium.Journalofbacteriology193,2027-2034(2011).

57. Brophy,J.A.N.etal.EngineeredintegrativeandconjugativeelementsforefficientandinducibleDNAtransfertoundomesticatedbacteria.NatureMicrobiology3,1043-1053(2018).

58. Bose,B.&Grossman,A.D.RegulationofHorizontalGeneTransferinBacillussubtilisbyActivationofaConservedSite-SpecificProtease.JournalofBacteriology193,22(2011).


https://doi.org/10.1101/537571

23

59. Miwa,Y.,Nakata,A.,Ogiwara,A.,Yamamoto,M.&Fujita,Y.Evaluationandcharacterizationofcatabolite-responsiveelements(cre)ofBacillussubtilis.Nucleicacidsresearch28,1206-1210(2000).

60. Kunjapur,A.M.&Prather,K.L.J.DevelopmentofavanillatebiosensorforthevanillinbiosynthesispathwayinE.coli.bioRxiv,375287(2018).

61. Ulrich,N.etal.ExperimentalstudiesaddressingthelongevityofBacillussubtilisspores–Thefirstdatafroma500-yearexperiment.PloSone13,e0208425(2018).

62. Nagler,K.,Setlow,P.,Li,Y.-Q.&Moeller,R.HighSalinityAlterstheGerminationBehaviorofBacillussubtilisSporeswithNutrientandNonnutrientGerminants.AppliedandEnvironmentalMicrobiology80,1314(2014).

63. Nicholson,W.L.,Munakata,N.,Horneck,G.,Melosh,H.J.&Setlow,P.ResistanceofBacillusEndosporestoExtremeTerrestrialandExtraterrestrialEnvironments.MicrobiologyandMolecularBiologyReviews64,548(2000).

64. Harrell,C.R.,Djonov,V.,Fellabaum,C.&Volarevic,V.RisksofUsingSterilizationbyGammaRadiation:TheOtherSideoftheCoin.Internationaljournalofmedicalsciences15,274-279(2018).

65. Kreyenschulte,D.,Krull,R.&Margaritis,A.RecentAdvancesinMicrobialBiopolymerProductionandPurification.CriticalReviewsinBiotechnology34,1-15(2014).

66. Bayer,T.S.&Samson,J.A.(GooglePatents,2015).67. Arkin,A.P.Awiseconsistency:engineeringbiologyforconformity,reliability,predictability.

CurrentOpinioninChemicalBiology17,893-901(2013).68. Mus,F.etal.SymbioticNitrogenFixationandtheChallengestoItsExtensiontoNonlegumes.

Appliedandenvironmentalmicrobiology82,3698-3710(2016).69. Dismukes,G.C.,Carrieri,D.,Bennette,N.,Ananyev,G.M.&Posewitz,M.C.Aquaticphototrophs:

efficientalternativestoland-basedcropsforbiofuels.Currentopinioninbiotechnology19,235-240(2008).

70. Sajid,M.,Nazal,M.K.,Ihsanullah,Baig,N.&Osman,A.M.Removalofheavymetalsandorganicpollutantsfromwaterusingdendriticpolymersbasedadsorbents:Acriticalreview.SeparationandPurificationTechnology191,400-423(2018).


https://doi.org/10.1101/537571

24

Figures:

Figure1: 3D printer design and development of the bioink components. a, The printhead

combinesthematerialandcellstreams;thisschematicshowsthecombinationofagaroseandB.subtilis

spores(bluecircles).Thenozzletemperatureandplatetemperaturesareheldconstantandtheambient

temperatureoftheprintchamberiscooled.Onceprinted,thesporescanbegerminated(greenovals)to

performtheirengineeredfunction.b,ThemodifiedMakerBotReplicator2Xhighlighting:1.theprinthead

design,2.modifiedpressurizedtank,3.coolingsubsystem,and4.electroniccircuitboard.Thelocations

ofthetemperaturesmeasuredintheprintheadareshown.DetailsareprovidedintheSupplementary

Information. c, Themelting temperature of 5% agarose is shown. The points represent data from 3


https://doi.org/10.1101/537571

25

independent replicates.d, The hardening (storagemodulus) of 5% agarosewhen the temperature is

shiftedfrom70°C(Tnozzle)to16°C(cooledprintchamber).Thepointsrepresentdatafrom3independent

replicates.Theinitialrateofgelation(dG’/dtast®0)iscalculatedandtheinsetshowsitsdependence

onthe%agarose.e,Theresponseofdifferentbacterialspeciesto20minutesofheatshockat75°C(red

lines). At t = 0, the bacteria are resuspended in fresh media and grown at their optimal growth

temperature (Methods). Each line corresponds toonegrowthexperiment,performed ingroupsover

threedays.BluelinesshowwhenB.subtilis168isfirstgrowninspore-inducingmedia(DSM)priortoheat

shock. Black linesshowgrowthwhenthere isnoheatshock.f,Comparisonofmaterialsproducedby

printingvegetativecellsorsporesofB.subtilis168constitutivelyexpressingGFP.Theblankbardoesnot

haveanycells.Scalebar,1mm.


https://doi.org/10.1101/537571

26

Figure2: 3Dprintingoflivingmaterials.a-d,Examplesofprintedobjectsareshownwhenthe5%

agarosematerialstreamisprintedalone. (e,f,g)ObjectsareshownwheresporesofGFP-expressingB.

subtilis168areprintedwiththeagarose.GFPisimagedbyilluminatingtheobjectwithbluelightfroma

SafeImager(Invitrogen,Carlsbad,CA).h,Anobjectwherethebottomhalfisprintedwithcellsandthe

top half without using two printheads (Methods). i, Comparison of printed parts with that of a

commerciallyavailableprinter(ProJet6000).Scalebars,5mm.


https://doi.org/10.1101/537571

27

Figure3: Cuttingtheobjectinducesgerminationofspores.a,Objectswereprintedwithsporesof

B.subtilisPY79constitutivelyexpressingGFP.b,Whenimagedimmediatelyaftercutting,germinatedcells

areonlypresentattheedge(Methods).c,Afterincubationfor10hoursinLBmedia,germinatedcellsare

presentonthenewsurface.Imagesrepresentativeofthreeexperimentsperformedondifferentdaysare

shown.Thegrayscalevalueobtainedisnormalizedbydividingthevaluesby255.Scalebars,3mm.


https://doi.org/10.1101/537571

28

Figure4: Inductionofgenome-encodedsensorsintheprintedmaterial.Sensorsweredesigned

forB.subtilisthatrespondto:(a)IPTG,(b)VanillicAcidand(c)Xylose(SupplementaryFigures30-32).The

bargraphs(left)showtheinductionofcells(1mMIPTG,1mMVanand1%Xylose,respectively)for90

minutesinLBmedia,measuredbyflowcytometry(Methods).Barswereprinted(2by3by25mm)with

sporesofB.subtiliscontainingthesensors.Thebarfluorescencewasimagedbeforeandafterinduction

for12hoursinLB.Thetimecoursesquantifythefluorescencesofthebars(right)asquantifiedfromthe

Images(Methods).Themeansandstandarddeviationsofthreeexperimentsareshown,performedon

differentdays.


https://doi.org/10.1101/537571

29

Figure5: Resilience of livingmaterials printedwithB. subtilis spores. a, Reconstitution of the

shapeoftheprinted(2x3x25mm)barsthathavebeendriedatroomtemperaturethenhydratedwith

water(Methods).b,Quantificationofthewidthofthebars.c,Afterdehydration,thebarswerestoredat


https://doi.org/10.1101/537571

30

ambienttemperatureonalabbenchfortheindicatedtimesandthenrehydrated.Thesporesinthebars

weregerminatedbyplacingtheminLBmedia.d,Barexposedtootherchallengingconditionssuchas

100%ethanol,highosmolarity(1.5MNaCl),UVlight(365and254nm)andg-radiation(2.6and15.75

kGy).ImagesoftheexpressionofGFPfromaconstitutivepromoterareshown(Methods).Themeansand

standarddeviationsofthreeexperimentsareshown,performedondifferentdays.Additionaltimepoints

andcontrolsareshowninSupplementaryFigure34-37.


https://doi.org/10.1101/537571

resilient living materials built by printing bacterial spores · living materials, where cells are...

Documents