Microreactor and probe chip for propane-eaters - UCLA

shamebagBiotechnology

Feb 22, 2013 (4 years and 1 month ago)

117 views

Glen

Summer 2009

Aerobes that eat propane


Rhodococcus

rhodochrous

PNKb1


ATCC 21197, 21198, $240, identified by vendor as propane
-
eater


Nocardia

paraffinicum


3.0 mL cell suspension


11
μ
mol n
-
propanol

in
40 minutes


Arthrobacter

CRL
-
60


Pseudomonas
fluorescens


Brevibacterium

butanicum


Mycobacterium
austroafricanum


ATCC 29678 $195, identified by vendor as propane
-
eater


Mixed culture


ATCC 21032 $240, identified by vendor as a mix of propane
-
eaters



Ashraf
, W.,
Mihdhir
, A., Murrell, C.
FEMS
Microbiol
.
Lett
.
1994
,
122
, 1
-
6.

Babu
, J.P., Brown, L.R.
Appl. Environ.
Microbiol
.

1984
,
48
, 2, 260
-
264.

www.atcc.org



Anaerobes that eat propane



Gram
-
positive (
i.e.
high
peptidoglycan

content in cell wall,
unlike most bacteria of human disease), spore
-
forming bacteria




Thermophilic
, marine species of genus
Desulfotomaculum
;
this genus also rots sealed, canned food




2 C
3
H
8

+ 5 SO
4
2
-

+ 4 H
+


5 H
2
S + 2 HCO
3
-


12
m
M

H
2
S formed in 3 months from bacteria floating in
one bottle of 100 mL medium and 56 mL gas


Kniemeyer

, O.
et al. Nature,

2007
,
449
, 898
-
902. www.atcc.org


ATCC sells 13 species of
Desulfotomaculum
,
and one of them (
D.
reducens
,
ATCC
BAA
-
1160, $195) has a completely known genome.


DSMZ (Germany) sells 29 species of
Desulfotomaculum


Neither vendor identifies any strain as a propane
-
oxidizing strain.

Anaerobic mechanism proposed

Hiro
, F.
J. Phys. Chem. B
,
2002
,
106
, 7688
-
7692.

Widdel
, F.,
Boetius
, A.,
Rabus
, R.
Prokaryotes,

2006
, 2, 1028

1049.

Becker, A., Fritz
-
Wolf, K.,
Kabsch
, J.,
Knappe
, S., Schulz, S., Wager, A.F.V.
Nature
Struct
. Biol.
,
1999
,
6,
969
-
975.


Can we see this step?

isopropylsuccinate

Enzyme
mech

based on a well
-
used motif

Enzymes that use the
glycyl

radical
mechamism

are difficult to isolate:



They need another enzyme to get activated


Exposure to O2 will break the peptide bond between
Gly

and
Cys

Attack on primary carbon proposed

During the 50 days of propane
consumption, some formation of
n
-
propylsuccinate

occurs.

n
-
propylsuccinate

Kniemeyer

, O.
et al. Nature,

2007
,
449
, 898
-
902. www.atcc.org



This step was newly proposed in
2007. It happens for only about 1/3
of the propane oxidation reactions



Primary C
-
H bond cleavage is more
difficult than secondary.

Evidence of attack primary C

Metabolites were extracted
in CH2Cl2 and
methylated
.


GC
-
MS was compared to
standard sample of n
-
propylsuccinate



We have GC
-
MS in the
shared instrument rooms.

Kniemeyer

, O.
et al. Nature,

2007
,
449
, 898
-
902


Planktonic

anaerobes, anchored aerobes


Anaerobes:
Kniemeyer

et al.

used
planktonic

Desulfotomaculum


But microbes live in sediments ( =>native state is anchored)


Anchored state might be easier to maintain over several days inside a
hollow
-
fiber membrane
microbioreactor

; Chung et al did this for 50 days


Slow
rxn

rate for anaerobes => cell maintenance is necessary



Aerobes:
Babu

and Brown used cell suspensions of
Nocardia

paraffinicum

(
planktonic
)


But microbes live in soil, sediments => native state is anchored


Propane oxidation occurs in minutes, so
microbioreactor

does not need
to maintain cells for long.


Anchoring to membrane increases cell density in reactor


higher
concentration of metabolites for better NMR detection


Anchoring to membrane with ~0.1
μ
m pores allows controlled flow of
nutrients, waste


Planktonic

microbioreactor


Six
microbioreactors

on one
PDMS
-
on
-
glass chip



Bacteria in a constantly flow
in a loop to prevent
biofilm

formation


Each flow loop has nine segments


Each segment is individually isolated for cleaning
during the experiment


Lysis

buffer is pumped from on
-
chip well to kill
and to detach
biofilm

cells


Extra growth medium is pumped from on
-
chip
well to re
-
dilute cells for control of population size.

Balagaddé
,
F.K.,You
, L., Hansen, C.L., Arnold, F.H., Quake, S.R.
Science,
2005
,
309
, 137
-
140.



Make the

Soft Lithography Master

1.
Design and print mask.


2.
Spin SU
-
9
photoresist

to Si
wafer


3.
Expose wafer through mask,
develop and bake
photoresist


4.
The wafer is now the master
template. Resolution ~ 20
μ
m

McDonald, J.C., Duffy, D.C., Anderson, D.T., Chiu,
D.T., Wu, H.,
Schueller
, O.J.A.,
Whitesides
, G.M.
Electrophoresis,
2000
,
21
, 27
-
40.


Make the PDMS layer on the master

McDonald, J.C., Duffy, D.C., Anderson, D.T., Chiu, D.T., Wu, H.,
Schueller
, O.J.A.,
Whitesides
,
G.M.
Electrophoresis,
2000
,
21
, 27
-
40.



Place newly cast PDMS layer onto glass
(“flat”).


Align PDMS and flat.


Using air plasma,


Make two layers


Top PDMS layer for air channels


Air pressure ≈ +5 psi


PDMS 4mm thick for channel stability



Bottom PDMS layer for fluid channels


PDMS 30
μ
m thick via spin
-
coat onto master,
2000 rpm, 30 s.


Channels 100
μ
m wide x 9
μ
m high



Valve = intersection of fluid and air channels



Unger, M.A., Chou, H
-
P.,
Thorsen
, T., Scherer, A., Quake, S.R.
Science,
2000
,
288
, 113
-
116.


Thorsen
, T.,
Maerkl
, S.J., Quake, S.R.
Science,
2005
,
298
, 580
-
584.


McDonald, J.C., Duffy, D.C., Anderson, D.T., Chiu, D.T., Wu, H.,
Schueller
, O.J.A.,
Whitesides
, G.M.
Electrophoresis,
2000
,
21
, 27
-
40.



The two layers are different types of PDMS: one layer contains free Si
-
H groups, the
other layer contains free vinyl (
ethenyl
) groups; the two bond covalently and become a
monolith.


General Electric RTV 615 silicone potting compound kit, 1 pint for $109.55 from
local vendor RS Hughes (Pacoima, CA)


Peristaltic pump made of PDMS
channel intersections

Add a drainage capillary for NMR detection



In lieu of six separate reactors, use only two.


Use extra space on chip for PDMS peristaltic pump


Pump cell suspension segment at a time to NMR detection device


Connect chip to NMR detector via 100
-
μ
m capillary

NMR
detection
device

Immobilization: denser cell growth


Membrane


Flat membrane


Hollow
-
fiber membrane



Free
microcarriers


Commercial designer particles



Packed bed


Foam


Polymer


Gel


Beads


Raschig

ring


Coke


Activated carbon



17 mL

Hollow
-
fiber membrane



Even
distribution of
cells




High surface
area / volume




Control of
nutrient and
waste flows



I would use one
1.1
-
mm fiber in
a glass tube
microbioreactor


Separation of (g), (l) nutrients possible

0.33 mm ID

0.66 mm OD

0.4
-
0.6
μ
m


pores

1.47 mm ID

1.96 mm OD

8 mm ID

Chung, B.H., Chang, H.N., Kim, I.H.
Enzyme Microb. Technol
.,
1987
,
9
, 344
-
349.

1.3 cm OD
microbioreactor

15 cm

1/16 “ ID x 0.025” OD
peristaltic pump tubing feeds
hollow
-
fiber membrane directly.

This is for seeding the fiber with
live cells, and for collecting
metabolites.

1/16 “ ID x 0.025” OD
peristaltic pump tubing takes
output to optical microscope and
to NMR. Tubing fits through
rubber septum with sealant.

Glass reactor contains a single
Romicon

XM50 hollow
-
fiber
tube, 1.1 mm in diameter.

Nutrients and dissolved propane
gas flow around outside of
hollow
-
fiber membrane, via
peristaltic pump and ¼” tubing.

9 mm OD

Optical microscopic observation to check for
escaping cells (cellular attachment)

Glass plate, 1” x 5”

Glass plate, 1” x 5”

5” length allows glass to overhang
microscope stage on both sides

PDMS sandwiched
in the middle here, in
the shape of one
microfluidic

channel, 100
μ
m
wide, 10
μ
m deep.

360
μ
m holes for metabolite throughput:
connections to 1/16” ID peristaltic pump
tubing made by
Nanoport

PEEK connectors
(IDEX Corp)

This hole on top plate

This hole on bottom plate

Output goes to NMR

Detachment of bacteria from bioreactor

Johansen, C.,
Falholt
, P., Gram, L.,.
Applied Environmental
Microbiology,

1997,
63
, 3724


3728.

Bockelmann
, U.,
Szewzyk
, U.,
Grohmann
, E.
Journal of
Microbiological Methods,
2003,
55,

201


211.

For use of live cells in the
stripline

probe,
cells detach from hollow
-
fiber membrane
walls with buffered mixture of
α
-
glucosidase
,
β
-
galactosidase
.



Commercial enzyme cocktail
Pectinex

Ultra
SP
-
L works, too, but may be more expensive.

NMR
detection
device

Strip of Cu instead of coil

van
Bentum
, P.J.M., Janssen, J.W.G.,
Kentgens
, A.P.M., Bart, J.,
Gardeniers
, J.G.E.,
Journal of Magnetic Resonance,
2007
, 189, 104
-
113.


A Dutch group uses a strip of copper instead of a coil to receive
and to send
rf

signals. Their
1
H frequency is 600
MHz.




Thin strip part of Cu is 35
μ
m thick, 1 mm long, 500
μ
m wide,
etched out of the Cu foil coating the Rogers 5870 PTFE material.

B1 homogeneity for pulses
perpendicular to B
0

field is
shown in yellow in this
Maxwell equations
simulation:

Microfluidic

channels may run here

(axial view)

Homogeneity might be better with
stripline

than with
microcoils
?

Axial cross
-
section of half of
a four
-
turn 3D coil. Sample
would go into the rectangle.

Axial cross
-
section of half of a
three
-
turn 2D coil (spiral).
Sample would go into the
rectangle.

van
Bentum
, P.J.M., Janssen, J.W.G.,
Kentgens
, A.P.M., Bart, J.,
Gardeniers
, J.G.E.,
Journal of
Magnetic Resonance,
2007
, 189, 104
-
113.

http
://
www.ndt
-
ed.org/EducationResources/CommunityCollege/MagParticle/Graphics/coil1.gif

Double
-
resonance
stripline

probe


van
Bentum
, P.J.M.,
et al. Journal of Magnetic Resonance,
2007
,
189
, 104
-
113.



Bart, J.,
Kolkman
, A.J.,
Oosthoek
-
de
Vries
, A.J., Koch, K.,
Nieuwland
, P.J., Janssen,
J.W.G., van
Bentum
, P.J.M.,
Ampt
, K.A.M.,
Rutjes
, F.P.J.T.,
Wijmenga
, S.S.,
Gardeniers
,
J.G.E.,
Kentgens
, A.P.M.
J. Am. Chem. Soc.
2009
,
132
, 14, 5014
-
5015.

microreactor

Stripline

is easier to build than coil


(sputter Cu or
anodically

bond Cu strip)


Tune,
match via
dielectric
plunger

5
-
mm
microcoil

versus
stripline

Human cerebrospinal fluid
using 256 scans, 6
-
μ
s 90


pulsewidth

on 5mm
commercial
microcoil

Human cerebrospinal fluid
using 4608 scans, 7
-
μ
s 90


pulsewidth

on
stripline


Bart, J.,
Kolkman
,
et al.

J.
Am. Chem. Soc.
2009
,
132
,
14, 5014
-
5015.

Ways to improve sensitivity


Bart, J.,
Kolkman
,
et al.

J. Am. Chem. Soc.
2009
,
132
, 14, 5014
-
5015 online supplement

Knapkiewicz
, P.,
Walczak
, R.,
Dziuban
, J.A.
Optica

Applicata
,
2007
,
37
, 65
-
72.

Olson, D.L, Peck, T.L., Webb, A.G.,
Magin
, R.L.,
Sweedler
, J.V.

Science
,
1995
,
270,
1967
-
1970.


Use nonconductive substrate instead of Si.


Nanolab

has anodic bonding machine


Forturan

glass: bond at 250

C, under 2kV



Flow two
microfluidic

channels along length
of
stripline

(one on each face of strip)

vs.

Forturan

glass

Si

Rogers 5870 can be ordered as a PTFE sandwiched between
two copper foil layers (35
μ
m)

A piece of Rogers 5870 with
copper foil cladding

Mill channel into dielectric surface (
fluoropolymer
):

100
μ
m wide, 10
μ
m deep

Thinnest capillary wall for 100
μ
m ID = 64
μ
m;




Height of homogeneous B
1
field of
stripline

≈ 50
μ
m

Therefore, sample must flow into shallow channel in dielectric.

Gershenfeld

et al
used capillary, but did not show homogeneity of B
1

Bentum

et al
used channel, and showed limited homogeneity height


Copper ground planes

Copper to become
the planar NMR
rf

strip

PTFE to become the
microfluidic

channel bed

Etch away copper on portions of one
piece, for
microfluidic

channel


Mill a channel:

100
μ
m wide, 10
μ
m deep


Place input and output where
convenient.


For example, if ends are
convenient, then glue glass strip
or PTFE strip to ends, and etch
360
μ
m hole in each for IDEX
Nanoport

capillary connections

Stripline
-
bearing Rogers
5780 piece will form lid to
this channel. Use PTFE
adhesive on PTFE parts
only, and press hard.

Channel volume on order of
tens of
nanoliters


Copper ground planes

Copper to become
the planar NMR
rf

strip

PTFE to become the
microfluidic

channel bed

Shorter piece of Rogers 5870

Longer piece of Rogers 5870

Mill channel into dielectric surface (
fluoropolymer
):

100
μ
m wide, 10
μ
m deep

Thinnest capillary wall for 100
μ
m ID = 64
μ
m;




Height of homogeneous B
1
field of
stripline

≈ 50
μ
m

Therefore, sample must flow into shallow channel in dielectric.

Channel volume on order of tens of
nanoliters

Affix
planktonic

microbioreactor

here.

Connect to NMR sample
channel via capillary
tubing

Affix Quaker
-
design
planktonic

microbioreactor

onto
dielectric, and cap NMR sample channel

Affix
planktonic

microbioreactor

here.

Connect to NMR sample
channel via capillary
tubing

Stripline
-
bearing Rogers
5780 piece will form lid to
this channel. Use PTFE
adhesive on PTFE parts
only, and press hard.

Single
-
cell imaging: no?

These are images of very large cell, the
Xenopus

oocyte
.


Even for a macroscopic cell, the resolution is low , and
the individual pixels are distinguishable.

This is the T1


weighted image.

Purea
, A., Neuberger, T., Webb, A.G.,
Concepts in Magnetic Resonance Part B:

Magnetic Resonance Engineering,
, 2004, 1, 7
-
14.

Can we see this step?

Can we see this step?

Widdel
, F.,
Boetius
, A.,
Rabus
, R.
Prokaryotes,

2006
, 2, 1028

1049

Another use for hollow
-
fiber

Butterfield DA, Bhattacharya D,
Daunert

S,
Bachas

L.
J
Membr

Sci

2001,
181
, 29

37

Charcosset
, C.
Biotechnology Advances,
2006
,
24
, 482

492.

Bouchard, L., Burt, S.R., Anwar, M.S.,
Kovtunov
, K.V.,
Koptyug
, I.V., Pines, A.
Science

2008
,
319,

442
-
445.



Binding enzymes to membranes for high
-
throughput catalysis can be random (
A
)



B

is far more desirable




Characterization of true bound orientation
is laborious, via enzyme kinetics assays




Microfluidic

MRI can show
B

/
A
ratio and
other orientation details, just as done for
hydrogenation, in a single experiment.

Monitor growth inside tube

Linton, E. A.,
Higton
, G.. Knowles, C. J., and Bunch, A. W.
Enzyme and Microbial
Technology,

1989,
11
, 238
-
288.

Hydrostatic pressure increases as growth blocks

pores in tube walls;


Pressure transducer at inlet can measure this.



Growth inside tube affects productivity

Total protein,
including fiber
walls, lumen, and
outside of fibers.

Protein on inside
surfaces of fiber
walls only.

Pressure indicates when
to switch to maintenance
medium.

Fiber walls will not retain
original pore size when
pressure > 1psi

Immobilized cells grow more densely

Pore size = 100
μm

Cuture

time = 30 days


Polymer disk dimensions:

2.3 cm diameter, 50 mm height


I do not have an easy way

of observing these cells via

NMR; they are stuck.

Polymer is a polystyrene developed by the
authors, prepared from monomers and Span
80 surfactant in 220 mL water at RT with
three orthogonal mixing paddles