of 52
1
Supporting Information for
Widespread detoxifying NO reductases impart a distinct isotopic
fingerprint on N
2
O under anoxia
Renée Z. Wang, Zachery R. Lonergan, Steven A. Wilbert, John M. Eiler, Dianne K.
Newman
Corresponding authors: eiler@caltech.edu, dkn@caltech.edu
This PDF file includes:
Supporting text
Figures S1 to S19
Tables S1 to S13
SI References
1
Supplementary Information Text for “Widespread detoxifying NO reductases impart a
distinct isotopic fingerprint on N
2
O under anoxia”
Table of Contents
1.
Medium and nitric oxide donors
2.
Strain generation
3.
N
2
O Screen
4.
Headspace sampling and N
2
O distillation
5.
Site Preference Measurements
5.1 Delta (δ) Notation and Definition of Site Preference (SP)
5.2 Correction to international reference frame
5.3 SP Measurement workflow
5.4 Shot noise error and limits of precision
5.5 Zero enrichment tests and instrument performance
5.6 Measurement consistency across instruments
5.7 Scrambling correction
6.
Isotopic measurement of DETA NONOate and nitrate substrate
7.
Rayleigh plots of NOR
-
only strains
8.
Proteomics
1. Medium and nitric oxide donors
Synthetic Cystic Fibrosis Media
-
Amended (SCFM
-
A) includes a 1x trace metal stock derived from
a 1000x stock. A 1000x solution of the trace metal stock (Trace element sol. SL
-
10; DSMZ) at a
total volume of 1000 mL comprised: 1) 10.00 mL of HCl (25%; 7.7 M)
; 2) 1.50 g of FeCl
2
x 4
H
2
O; 3) 70.00 mg of ZnCl
2
; 4) 100.00 mg of MnCl
2
x 4 H
2
O; 5) 6.00 mg of H
3
BO
3
; 5) 190.00 mg
of CoCl
2
x 6 H
2
O; 6) 2.00 mg of CuCl
2
x 2 H
2
O; 7) 24.00 mg of NiCl
2
x 6 H
2
O; 8) 36.00 mg of
Na
2
MoO
4
x 2 H
2
O;
9) 990.00 mL of distilled water. The small molecule NO donor DETA
NONOate (C
4
H
13
N
5
O
2
, #82120 Cayman Chemical Company) decays following first order kinetics
in a pH
-
dependent manner to release two moles of NO per mole of DETA NONOate (half
-
life of
20 hours at 37°C and pH 7.4; see Fig. S4).
2. Strain generation
Strains were generated via allelic exchange. Briefly, ~1 kb fragments surrounding the gene
of interest were amplified by PCR and Gibson cloned into pMQ30
(1)
. Deletion constructs were
introduced into
Pa
via triparental conjugation, and
E. coli
plasmid and helper strains were selected
against on VBMM containing 50 μg/ml gentamicin
(2)
. Resulting Gent
R
Pa
cells were plated on
10% sucrose LB agar to isolate recombinants and screened via PCR. See Table S5 for primers
used. Another strain,
ΔnosZΔnorBC
, was also constructed but it did not grow appreciably in the
anaerobic, batch culture growth condition (Fig. S2); therefore, its SP was not measured.
3. N
2
O Screen
All strains were first grown to a high density (OD
600
~ 3
-
4) from glycerol freezer stocks in
aerobic pre
-
growths (25 mL SCFM
-
A, 250 rpm shaking for 16 hours at 37°C). Cells were then
pelleted and fully re
-
suspended into 25 mL of fresh media in sealed, glass 18 x 150 mm Balch
tubes. The headspace was then pur
ged with N
2
gas to establish anoxia, and 500
M DETA
NONOate was added. Balch tubes were incubated statically for 24 hours at 37°C. The headspace
2
was then sampled on the vacuum line and distilled to concentrate N
2
O and CO
2
in a preliminary
distillation. See Table S2 for results.
4. Headspace sampling and N
2
O distillation
N
2
O was distilled from the headspace samples on an ultra
-
torr vacuum line prior to isotopic
analysis (Fig. S14). First, the sample was expanded onto the left side of the line (Step 1); higher
pressure samples were sampled by taking multiple aliquots while lo
wer pressure samples were
fully exposed to the line. Next, non
-
condensables (i.e. N
2
, Ar) were removed by passing the sample
over a trap in liquid nitrogen (LN
2
, T2 in Fig. S14) and removing the residual headspace. Then,
the sample was passed back and fo
rth
over the ascarite tube and the ethanol / dry ice slurry trap
(T3, Fig. S14) to remove CO
2
and H
2
O. Each pass lasted four minutes. The sample was isolated
from the vacuum and the directionality of the sample flow was determined by either submerging
T1 (clockwise flow) or T2 (counter
-
clockwise flow) in
LN
2
. The ascarite tube was re
-
made roughly
every six samples; it consists of a ≈10” length pyrex tube of ⅜” gauge containing sodium
hydroxide (Ascarite II CO
2
Absorbent, Thomas Scientific) and sealed with quart
z wool on both
ends. The ethanol / dry ice trap was a slurry of 100% ethanol (v/v) mixed with dry ice (solid CO
2
).
Finally, the sample was passed over the ethanol / dry ice slurry for a final time and flame
-
sealed
into a pyrex glass finger until isotopic analysis.
Two vacuum distillation blanks (DistillationBlank
-
1, DistillationBlank
-
2) and a no
-
cells
vacuum flask blank (FlaskBlank
-
1) were measured to test if the distillation process causes
significant isotopic fractionation (Table S8). A total mixture of 640 μmol C
O
2
and 290 μmol N
2
O
were expanded to a total volume of 127 cc on the vacuum line, then equilibrated with a pyrex
finger of ~5 cc containing room air and ~0.1 mL DI water. Two aliquots of this mixture were taken
as mock samples (DistillationBlank
-
1, Distill
ationBlank
-
2). The no
-
cells vacuum flask blank
(FlaskBlank
-
1) was prepared as a batch culture, but after the headspace was purged with N
2
gas,
N
2
O from the reference tank was injected into the flask. This flask was then incubated at 37°C for
~12 hours and sampled at end
-
exponential phase (~12 hours). The Distillation blanks showed little
difference from the original N
2
O gas (roughly 0.1 ± 0.5‰ difference), indicating that the
distillation process does not significantly fractionate our target gas. The Flask
Blank showed a
difference of
-
2.25 ± 0.90‰ in δ
18
O; this may have been caused by exchange of O isotopes between
the incubated N
2
O gas and H
2
O
therefore our study relies on interpretation of the N isotopes
instead.
5. Site Preference measurements
5.1 Delta (δ) Notation and Definition of Site Preference (SP)
All isotopic measurements in this study are reported in the delta notation (δ) in units of per
mille (‰) where:
!"
=
%
!"
$
#$%
!"
$
&'(
1
(
×
1000
Equation S1
Where
15
R
is the ratio of
15
N/
14
N in the sample (“sam”) or reference (“ref”). All values here
are reported to the international reference of Air for nitrogen.
Site Preference (δ
15
N
SP
or “SP” in this study) is defined as the relative, intramolecular
enrichment of the rare, stable isotope
15
N for the central vs. terminal nitrogen in the linear,
3
asymmetrical N
2
O molecule. To be consistent with prior work, we use the designations as defined
by
(3,
4)
where the terminal nitrogen is labeled
, and the central nitrogen is labeled
. In this
convention, the
15
R ratios for each site is defined as:
!"
%
=
[
!)
'
!"
'
!*
(
]
[
!)
'
!)
'
!*
(
]
Equation S2
!"
*
=
[
!"
'
!)
'
!*
(
]
[
!)
'
!)
'
!*
(
]
Equation S3
Therefore, in delta notation:
!"
%
=
%
!"
$
+
#$%
!"
$
+
&'(
1
(
×
1000
Equation S4
!"
*
=
%
!"
$
,
#$%
!"
$
,
&'(
1
(
×
1000
Equation S5
Site Preference is as defined as in
(4)
:
푆푃
!"
%
!"
*
Equation S7
5.2 Correction to international reference frame
The working reference gas (“Caltech Ref Gas”) used in this study was previously
characterized relative to the international working standards for nitrogen and oxygen isotopes
(Air and VSMOW respectively) by Tokyo Tech
(5,
6)
. Values are listed in Table S9 below.
5.3 SP
Measurement Workflow
SP measurements were performed on two versions of the Thermo Scientific Ultra High
-
Resolution Isotope Ratio Mass Spectrometer (HR
-
IRMS), the ‘Prototype Ultra’
(7)
and the
‘Production Ultra.’ Two measurements were performed on each sample
the first at Mass 30 and
31 for δ
15
N
, and the second at Mass 44, 45 and 46 for δ
15
N
bulk
and δ
18
O. All measurements were
corrected for background (“Johnson”) noise
(7)
. Background correction was done before and after
measurement on
-
peak to adjust for any pressure
-
related intensity changes, or other instrument
changes over the course of the measurement. Gas injection pressures were calibrated to the main
peak (Mass 30 or
44), and all sample measurements were bracketed by the reference gas.
A Mass 45 foot correction was done to correct for a
13
C
16
O
2
‘foot’ that overlaps with the
14
N
15
N
16
O /
15
N
14
N
16
O measurement ‘shoulder.’ This ‘foot’ is present in both reference and sample
gases; reference gas was obtained through Matheson Gas at Ultra High Purity (99.99%). The
correction was done by calculating a pressure
-
varying ratio of the foot vs. shoulder, th
en applying
this pressure
-
varying ratio over the course of the measurement block. Two foot correction
4
observations bracketing the measurement on
-
peak on the ‘shoulder’ were done to account for any
pressure
-
related intensity changes over the course of the measurement. Over the observation
period, both the foot and shoulder signal will decay exponentially wi
th pressure, so both signals
were fit with equations for exponential decay:
+,,-
=
+,,-
퐸푋푃
(
+,,-
)
Equation S8
.
/
,01234
=
.
/
,01234
퐸푋푃
(
.
/
,01234
)
Equation S9
Where
I
= signal intensity,
t
= time, and
a
and
b
are fitted constants. We can then take the
ratio of both equations for the correction:
5
(--.
5
#
/
-012'&
=
6
(--.
6
#
/
-012'&
퐸푋푃
[
(
+,,-
.
/
,01234
)
]
Equation S10
Then this correction can be applied to the raw Mass 45 signal for the corrected Mass 45
signal:
4
5
7,44
=
?
1
6
(--.
6
#
/
-012'&
퐸푋푃
[
(
+,,-
.
/
,01234
)
]
@
4
5
468
Equation S11
This correction generally caused δ
15
N
bulk
to become more negative by roughly 1‰.
5.4 Shot noise error and limits of precision
In addition to background noise, shot noise is another inherent limit of isotope ratio
measurements
(7
9)
. Therefore, for each measurement we calculated the shot noise error and
compared it to the actual observed standard deviation of the measurement to see how close we
approach shot noise limits. Figure S15 shows calculated shot noise vs. observed standard d
eviation
for all measurements made for this study (samples, zero
-
enrichments, sample reruns;
n
= 79) across
both Prototype and Production Ultras over six experimental sessions over two years. Linear
regression across all points gives a slope close to 1 (
m
= 1.14 ± 0.04), indicating that overall
measurements are approaching the shot noise error. Median calculated shot noise across all
measurements (0.49‰, Fig. S15) is less than the median measured std. dev. (0.61‰, Fig. S15),
indicating that, overall, measu
r
ements were reaching shot noise limits.
5.5 Zero enrichment tests and instrument performance
‘Zero enrichment’ tests where the reference gas is measured as a sample against itself were
regularly performed over the course of the study to ensure that pressure balance for the sample and
reference gas bellows were correctly calibrated
i.e., if the b
ellows are correctly pressure
calibrated, we would expect each measurement (δ
15
N
bulk
, δ
18
O and δ
15
N
ɑ
) to give a value of 0‰
within uncertainty (1 s.d.). Fig. S16
shows the result of eight zero enrichment tests for A) δ
15
N
bulk
,
C) δ
18
O and E) δ
15
N
ɑ
run on the Prototype (pink circles) and Production Ultra (blue circles).
Results are largely 0‰ within uncertainty (1 s.d.) with the exception of δ
15
N
bulk
and δ
18
O for Jun.
5
2021, and δ
15
N
bulk
for Mar. 2023. However, this offset is on the order of 0.1‰, which is within the
uncertainty of our shot noise error and is therefore likely due to the inherent limits of precision in
our measurement. In addition, zero enrichments were run over a range of
bellow pressures to gauge
instrument performance across sample size (Fig. S16 B,D,F). A proxy for bellow pressure is ion
beam intensity, since increased sample volume causes increased ion beam intensity. Calculating a
linear correlation
(not shown on figure) with minor ion intensity as the independent variable and
δ values as the dependent variable gives adjusted R
2
values of 0.09, 0.14 and
-
0.12 for δ
15
N
bulk
,
δ
18
O and δ
15
N
ɑ
respectively (calculation performed using R Statistical Software (v4.1.0; R Core
Team 2021,
(10)
) [call:
lm
()]. Therefore, there is no correlation between minor ion intensity and δ
values, indicating that our measurement method is accurate across a range of sample sizes, bellow
pressures, and signal intensities.
5.6 Measurement consistency across instruments
Two samples, 0225 and 0230, were measured on both the Prototype and Production Ultras
to
gauge measurement consistency across instruments (Fig. S17). Samples were first measured in
April 2022 on the Prototype Ultra. They were then removed from the sample bellow by freezing
into a small glass finger with a finger
-
twist valve using liquid nitrog
en (LN
2
). Samples were then
taken to the vacuum line, frozen into glass break
-
seals using LN
2
, and stored in break
-
seals until
further measurement. Samples were then re
-
measured on the Production Ultra on Dec. 2022.
Measurements of 0225 and 0230 on both Ul
tras give the same value within measurement
uncertainty (1 s.d.) vs. the reference gas for δ
15
N
bulk
, δ
18
O and δ
15
N
ɑ
, and all measurements
approach the shot noise limit where std. dev. to shot noise ratio is 1 (Fig. S17). Standard deviation
for both 0225 and 0230 is lower on the Prototype Ultra because there was more total sample in the
first measurement (i.e. some samp
le was consumed during the first measurement).
5.7 Scrambling Correction
Raw SP measurements must be corrected for ‘scrambling,’ a measurement artifact in which
the ionization process in a gas source mass spectrometer “scrambles” all isotopologues of N
2
O,
causing the inner (
)
nitrogen and the outer (
)
nitrogen appear to be switched
(11)
. There exist
multiple strategies for scrambling corrections
(12
15)
, and there are largely three levels of
complexity that the correction can be performed at: i) A single
-
factor correction (
) that assumes
the scrambling behavior of
14
N
15
N
16
O and
15
N
14
N
16
O are equal, and that the contribution of
17
O is
negligible at Mass 31
(4,
16)
; ii) A two
-
factor correction (
and
) that accounts for the difference
in scrambling between
14
N
15
N
16
O and
15
N
14
N
16
O, and assumes that
17
O follows a mass
-
dependent
relationship with
18
O
(17,
18)
; and iii) A nine
-
factor correction that accounts for differences in
scrambling between
14
N
15
N
16
O,
15
N
14
N
16
O,
15
N
15
N
16
O,
14
N
14
N
17
O,
14
N
15
N
17
O, and
15
N
14
N
17
O
(15)
.
We used the single
-
factor correction following
(3,
4)
because the nine
-
factor scrambling
correction
(15)
requires measurement of up to nine external reference gases, which we did not
have, and because we believe the scrambling effects of
15
N
15
N
16
O,
14
N
14
N
17
O,
14
N
15
N
17
O, and
15
N
14
N
17
O are negligible at the level of precision needed for this study
i.e. the variations in SP
between NOR and Fhp are on the order of 10‰. We did not use the two
-
factor correction following
(17,
18)
because we were able to mass resolve
17
O directly, and because that method is optimized
for continuous
-
flow SP measurements.
We measured two replicates (RM5_1 and RM5_2) of external reference gas RM5
(14)
on
the Production Ultra to calculate the scrambling factor (Table S10). Replicates were measured
6
three months apart, and RM5_2 was measured at a lower sample amount. RM5 was used because
it has a large, ~10‰ difference in δ
15
N between δ
15
N
bulk
and δ
15
N
. We measured δ
15
N
bulk
within
uncertainty for RM5_2, and close to within uncertainty for RM5_1 (Table S10). We consistently
measured the mean δ
15
N
value to be more depleted by roughly 1‰, though all measured δ
15
N
values overlapped with the reported δ
15
N
value within uncertainty. This caused the SP values of
RM5_1 and RM5_2 to be ‘compressed’ towards 0‰ compared to its reported value. δ
18
O values
for RM5_1 and RM5_2 were measured to be their reported values within uncertainty (Table S10).
We then followed
(4)
to calculate
= 0.04 ± 0.08 for RM5_1, and
= 0.05 ± 0.11 for
RM5_2. We therefore use an average
value of 0.045 ± 0.136 for samples measured on the
Production Ultra.
(5,
6)
used the Prototype Ultra and measured samples in similar tuning
conditions as used in this study; they used a one
-
factor correction of 0.110 ± 0.002. We therefore
used
= 0.110 ± 0.002 for samples measured on the Prototype Ultra.
was likely lower on the
Production Ultra because it has a lower baseline source pressure than the Prototype Ultra (3x10
-
10
vs. 9x10
-
8
mbar respectively).
The one
-
factor scrambling correction was performed as follows; an example correction is
shown for one measurement of iFhp (Table S11). First, the sample is measured vs. Caltech Ref
Gas. Next, values are corrected to international standards (AIR for N, VSMO
W for O) using values
reported by Tokyo Tech values (Table S11). Finally, the scrambling
-
adjusted
15
R
value (
15
R
adj
)
is calculated from the measured value (
15
R
meas
) and the measured bulk value (
R
bulk
meas
):
629
%
=
:36.
%
2
:36.
;01<
2
+
1
Equation S12
The final reported values, with scrambling correction and reported vs. AIR, are shown in
the rightmost column of Table S11.
We checked our scrambling
-
corrected values against previously reported
in vitro
values
for NOR.
Δ
nosZ
Δ
fhp
, which only has NOR, was corrected using
= 0.110 and iNOR was corrected
using
= 0.045. All corrected values overlap with previous
in vitro
measurements of a NOR
enzyme purified from
Paracoccus denitrificans
ATCC 35512
(19)
(Fig. S18). In addition, as an
internal check, the two samples measured on both the Production and Prototype Ultras (0225 and
0230, Fig. S17) gave similar values, implying that the scrambling factors are similar on both
instruments. Indeed,
= 0.045 ± 0.136 for the Production Ultra and
= 0.110 ± 0.002 are similar
within uncertainty.
6. Isotopic measurement of DETA NONOate and nitrate substrates
We calculated the average δ
15
N of the initial NO reactant used in the suspension assays by
measuring the difference in δ
15
N between the full and decomposed NO
-
donor, DETA NONOate
(#82120, Cayman Chemical Company). DETA NONOate (C
4
H
13
N
5
O
2
) is a pH
-
dependent NO
-
donor that decays following first order kinetics to release two moles of NO per mole of DETA
NONOate. At pH 7.4, it has a half
-
life of 20 hours at 37°C and a half
-
life of 56 hours at 22
-
25°C.
pH was adjusted using NaOH and HCl, then
the sample was prep
ared into 4x6 mm pressed tin
capsules (Costech Analytical Technologies) for analysis. The δ
15
N of the full molecule, which has
five nitrogens, and the decayed molecule, which was three nitrogens, was measured on a Delta
-
V
Advantage with Gas Bench and Costech elemental analyzer. Before measurement, the instrument
was tuned with an internal standard
to ensure instrument sensitivity and linearity, and to ensure
7
correct measurement mass position. Three analytical replicates of each sample were measured. All
samples were bracketed at the beginning and end of the run by a suite of external isotope standards
(Urea δ
15
N = 0.0‰; Acetanilide δ
15
N = 19.56 ± 0.03‰; all reported vs AIR), tin capsule blanks,
and NaOH and HCl blanks. After correcting for blanks, measured δ
15
N values were then corrected
to reported values vs. AIR using the external Urea and Acetanilide standards. On average, the
correction decreased the measur
ed δ
15
N values by 0.2‰.
We then calculated the average δ
15
N of the released NO molecules by mass balance using
the equation:
2
!"
=
'
=
5
!"
"
'
3
!"
>
'
Equation S13
Where δ
15
N
2N
, δ
15
N
5N
, and δ
15
N
3N
refer to the average δ
15
N of the released NO
molecules,
the full DETA NONOate molecule, and the decayed DETA NONOate molecule respectively. All
results are reported in Table S12.
δ
15
N values of the nitrate used in this study were measured in
the same way, except pH was not adjusted (Table S7).
Non
-
WT
Pa
strains were incubated with only nitrate or DETA NONOate as the NO source
to determine if the δ
15
N signal of the NO substrate was inherited in δ
15
N
bulk
(Fig. S8).
A. baumannii
,
iNOR, iFhp and
S. aureus
were all grown as suspension assays and showed δ
15
N
bulk
values of
-
91.0 ± 6.5‰ (mean ± s.d.). Though strains were incubated anoxically with both DETA NONOate
and nitrate, NO was only sourced from DETA NONOate because nitrate could not be reduced to
NO
i.e. native denitrification pathway was deleted in iNOR
and iFhp and does not exist in wild
-
type
A. baumannii
and
S. aureus
strains. Strains genetically inactivated for
nosZ
(
nosZ
) and
nosZ
along with
fhp
(
nosZ
fhp
)
strains were grown as batch culture
s and incubated anoxically with
only nitrate and showed δ
15
N
bulk
values of
-
27.4 ± 1.4‰. This implies a δ
15
N fractionation factor
of
-
27.8 ± 1.90‰ for nitrate to N
2
O, and
-
68.1 ± 6.5‰ for DETA NONOate to N
2
O.
7. Rayleigh plots of NOR
-
only strains
Rayleigh plots of two strains with only NOR, Δ
nosZ
Δ
fhp
and iNOR, were made to test if
the variation in SP followed a Rayleigh distillation relationship
(20)
. Δ
nosZ
Δ
fhp
was grown in
batch culture denitrifying conditions while iNOR was grown as a suspension assay.
For Δ
nosZ
Δ
fhp
, the fraction of nitrate consumed was calculated two ways
from the total
amount of N in each culture aliquot as measured on the Delta
-
V Advantage (
f
nitrate
), or from the
moles of N
2
O distilled from the headspace as measured in the direct
-
injection bellows on the
Prototype or Production Ultra (
f
N2O
). After each sampling time point for Δ
nosZ
Δ
fhp
where N
2
O
measurement was
performed (late
-
exponential and late
-
stationary), a roughly 5 mL aliquot of the
liquid culture was taken and immediately flash frozen in liquid nitrogen. Aliquots were then kept
frozen at
-
80°C until isotopic analysis. In addition, for each sampling batch,
an aliquot of KNO
3
,
SCFM
-
A media, and DI water were flash frozen and stored as well. When samples were ready for
analysis on the Delta
-
V Advantage with Gas Bench and Costech elemental analyzer as described
above, all flash frozen aliquots were thawed at r
oom temperature and aliquots were pipetted in
triplicate into individual 5x9 mm pressed tin capsules (Costech Analytical Technologies) and left
to dry overnight. Like the measurements described in Section 6 above, raw measurements were
corrected for tin ca
psule blanks using Urea and Acetanilide standards. Moles of total nitrogen were
calculated based on the total peak area of each sample. This analysis cannot distinguish between
N sourced from KNO
3
or SCFM
-
A, though there are roughly four times more moles o
f N from
8
KNO
3
vs. SCFM
-
A (Table S7). The fraction of total N remaining was calculated by dividing the
total N measured by the total N (KNO
3
and SCFM
-
A) initially added. This is referred to as
f
nitrate
.
We then calculated
f
N2O
for Δ
nosZ
Δ
fhp
:
'
=
(
=
1
=
@
3
4
5
@
35
6
7
Equation S14a
Where
n
N2O
are the moles of N
2
O produced and
n
NO3
-
are the moles of nitrate initially added.
N
2
O pressure was recorded in the sample bellows of the Ultra IRMS before each analysis, and
moles of gas were calculated using the ideal gas law. Following similar studies, this equation
assumes that every mole of nitrate that is taken up by the denitrificat
ion pathway results in two
moles of N
2
O
(21)
.
f
N2O
was calculated in a similar manner for iNOR, except moles of N
2
O produced were
compared to moles of DETA NONOate (
n
DETA
) added.
'
=
(
=
1
@
3
4
5
=
@
89:;
Equation S14b
We then constructed SP Rayleigh curves
(20)
following
(21,
22)
, where SP is plotted
against
-
(f*lnf)/(1
-
f), where
f
is the fraction of the remaining substrate. For Δ
nosZ
Δ
fhp
, two plots
were made using the two different ways that
f
was calculated (
f
nitrate
or
f
N2O
from Eqn. S14a
). For
iNOR, plot was only made using
f
N2O
calculated from Eqn. S14b
(Fig. S5).
Fitted values for
m
and
b
gave very large uncertainties and were of low confidence (Fig.
S5). In particular, fitted values using
f
N2O
had extremely large uncertainties due to the narrow range
of the x
-
axis
i.e. nitrate was given at saturating conditions so the nitrate pool was not very
depleted, resulting in
f
N2O
values around 0.99.
f
nitrate
varied over a larger range, likely because not
all the nitrate consumed ended up as N
2
O (i.e. due to assimilatory nitrate processes, or from loss
along the denitrific
ation pathway), so the approach for calculating
f
in Eqn. S14a likely gives an
overestimate. However, overall no correlation of SP with
f
is seen in the NOR
-
only strains,
nosZ
fhp
and iNOR.
9. Proteomics
For S
-
Trap
TM
digestion, pellets were resuspended in lysis buffer (5% SDS, 25 mM TEAB
pH 8.5) and sonicated for lysis. MgCl
2
was added to (2 mM final) concentration and incubated at
room temperature for 5 min. Samples were centrifuged for 10 min at 13,000 xg. Samples were
then reduced with TCEP (5mM final), and samples were alkylated by addition of MMTS (20 mM
final). Samples w
ere acidified with phosphoric acid (2.5% final concentration), mixed with
binding/wash buffer (100 mM TEAB, 90% methanol), and applied to the S
-
Trap column. Samples
were centrifuged at 4,000 xg for 30 s and washed 3 times with binding/wash buffer. S
-
Trap w
as
centrifuged for 1 min at 4,000 xg to fully remove binding/wash buffer. S
-
Trap was transferred to
1.7 mL microcentrifuge tube. Digestion buffer (1 μg trypsin/10 μg sample weight in 50 mM
TEAB) was added to the S
-
Trap, and the tubes were loosely capped an
d incubated for 1 h at 47°C.
40 ul of elution buffer 1 (50 mM TEAB in water) was applied to S
-
Trap and centrifuged for 1 min
at 4,000 xg. Elution and centrifugation was repeated with elution buffer 2 (0.2% formic acid in
9
water), and elution buffer 3 (50% acetonitrile in water). Eluted peptides were pooled, dried, and
resuspended in 0.2% formic acid for analysis.
The Q Exactive HF was operated in data
-
dependent mode with Tune (version 2.8 SP1 build
2806) instrument control software. Spray voltage was set to 1.6 kV, S
-
lens RF level at 50, and
heated capillary at 275°C. Full scan resolution was set to 60,000 at
m
/
z
200. Automatic gain control
target was 3 × 10
6
with a maximum injection time of 15 ms. Mass range was set to 375
1500
m
/
z
and charge state inclusion set to select precursors of charge state 2
5 for DDA analysis. For data
-
dependent ms2 scans, the loop count was 12, AGC target was set at 1 × 10
5
, intensity threshold
was kept at 1 × 10
5
, and dynamic exclusion set to exclude precursors after one time for 45 seconds.
Isolation width was set at 1.2
m
/
z
and a fixed first mass of 100 was used. Normalized collision
energy was set at 28. Peptide match was set to off, and isotope exclusion was on. Data acquisition
was controlled by Xcalibur (4.0.27.42), with ms1 data acquisition in profile mode and ms2 data
acquisition in centroid mode.
Data analysis was performed using Thermo Proteome Discoverer 2.5 (Thermo Fisher
Scientific, San Jose, CA) with a SEQUEST algorithm (PMID 24226387). The data was searched
against the
Pseudomonas aeruginosa
UCBPP
-
PA14 proteome (UP000002438) acquired from
UniProtKB (PMID: 36408920) in 2022
-
2
-
09. The ms1 matching tolerance was 20 ppm, and the
ms2 tolerance was 0.02 Da. Carbamidomethyl (+57.021 Da) on cysteine was set as static
modification, and Oxidation (+15.
995 Da) on methionine was set as dynamic modificatio
n.
Acetylation (+42.011 Da), Met
-
loss (
-
131.040 Da), and Met
-
loss + Acetylation (
-
89.030 Da) at
protein N
-
terminus were also set as dynamic modifications. A maximum of 2 miss
-
cleavages were
allowed in the search. A concatenated target decoy
-
based percolato
r was utilized to control the
false discovery rate. The q
-
value cutoff was set as 0.05. Protein abundances were reported using
ms1 feature
-
based label
-
free quantitation. The median abundance for each sample was normalized
to the same value.
10
SUPPLEMENTARY FIGURES
11
Fig. S1. Phylogeny of Fhp in
Bacteria.
(A)
Phylogram of annotated Fhp/Hmp amino acid sequences in the NCBI database; phylogram
was curated to show a representative group of bacteria. Strains in green were measured for SP in
this study. Fhp/Hmp from
S. aureus
shows 31.6% sequence similarity to Fhp from
P. aeruginosa
,
while Fhp from
A. baumannii
shows 98.5% similarity.
(B)
Tree showing abundance of Fhp, NorB
and NorC across Bacteria at the Phylum level, annotated in AnnoTree
(23)
and visualized using
the interactive tree of life (iTOL). Search parameters for AnnoTree were: % identity: 30; E value:
0.00001; % subject alignment: 70; % query alignment: 70.
12
Fig. S2. Growth Curves.
(A)
Growth curves measured by OD
600
for WT
Pa
,
nosZ
, and
nosZ
fhp
grown in batch culture,
denitrifying conditions with headspace sampling times for SP measurements (end
-
exponential,
red; end
-
stationary, yellow). Data points are mean ± standard deviation (
n
=6).
(B)
Growth curves
norBC
nosZ
strain in SCFM (black) and SCFM
-
A (white) media. Data points are mean ±
standard deviation (
n
=3).
13
Fig. S3. Construction of iNOR and iFhp strains.
(A)
Growth curve measured by OD500 of WT
Pa
(black),
nor
chromosomally complemented
with
nor
under a rhamnose
-
inducible promoter (
att
:
mTn7
(GentR,
norCBD
); blue) and
fhp
nor
nosZ
(purple) with rhamnose
-
inducible
nor
grown anaerobically in Lysogeny Broth
(LB) media with 40 mM potassium nitrate alone (upper panel) or 40 mM nitrate and 305
M
rhamnose (lower panel). Data points are mean ± standard deviation (
n
=3).
(B)
WT
Pa
(black),
fhp
chromosomally complemented with
fhp
under a rhamnose
-
inducible promoter
(
att
:
mTn7
(GentR,
fhp
); pink) and
fhp
nor
nosZ
(yellow) with rhamnose
-
inducible
fhp
were
grown aerobically in LB media with 500
M DETA NONOate alone (upper panel) or 500
M
DETA NONOate and 305
M rhamnose (
lower panel). Data points are mean ± standard deviation
(
n
=4).