Adapting your laboratory microplate reader for bacterial endotoxin testing

1
00:00:02.250 --> 00:00:04.051
Hello, my name is Timothy Francis.

2
00:00:04.110 --> 00:00:09.541
I'm the senior technical specialist for Fujifilm Wako Chemicals USA Corporation,

3
00:00:10.110 --> 00:00:15.150
and I'm excited to be providing this portion of this presentation on

4
00:00:15.255 --> 00:00:20.070
adapting your laboratory microplate reader for bacterial endotoxin

5
00:00:20.130 --> 00:00:26.100
testing with molecular devices. I'm going to be presenting

6
00:00:26.161 --> 00:00:32.131
to you an overview of the bacterial endotoxin test, as well as some of the studies

7
00:00:32.220 --> 00:00:36.481
we've done with the reader in the performance with our reagent.

8
00:00:39.360 --> 00:00:43.140
First, I'd like to talk about what the bacterial endotoxin test is.

9
00:00:44.100 --> 00:00:49.845
Here we see the USP85, the bacterial endotoxin test is the title of this

10
00:00:49.890 --> 00:00:55.620
slide. USP Chapter 85 outlines for us the bacterial endotoxin test

11
00:00:55.651 --> 00:01:00.720
with what is called the Limulus amebocyte lysate reagent, which is what our company

12
00:01:00.780 --> 00:01:06.390
manufactures. There are three techniques listed in USP85 for the

13
00:01:06.450 --> 00:01:11.970
bacterial endotoxin test. Now, this chapter is also harmonized with the European

14
00:01:12.030 --> 00:01:14.910
Pharmacopeia and Japanese Pharmacopeias as well.

15
00:01:15.705 --> 00:01:18.900
The first methodology is called the gel clot technique.

16
00:01:20.250 --> 00:01:25.711
This measures endotoxin by the presence of gel formation after cleavage of peptide

17
00:01:25.770 --> 00:01:30.240
bonds within coagulogen to yield coagulin.

18
00:01:31.140 --> 00:01:36.705
You can see the LAL cascade that I've illustrated here, showing the

19
00:01:36.840 --> 00:01:42.780
reaction that takes place with these protein factors when endotoxin

20
00:01:42.810 --> 00:01:48.601
is introduced. So the gel clot technique is literally just reacting

21
00:01:48.780 --> 00:01:54.180
LAL reagent with a sample in a test tube after incubation of one hour.

22
00:01:55.080 --> 00:02:00.480
At that point, the test tube is inverted and a positive test is recorded if

23
00:02:00.570 --> 00:02:06.450
that solution is clotted and remains at the top of the test tube.

24
00:02:06.510 --> 00:02:12.465
At that point, we would say that that sample has endotoxin, at least at the labeled

25
00:02:12.540 --> 00:02:17.310
sensitivity of the reagent. You can see this as a qualitative method.

26
00:02:17.370 --> 00:02:22.800
It only gives a positive or negative result at a specific sensitivity.

27
00:02:22.815 --> 00:02:28.005
The turbidimetric and chromogenic techniques, however, are quantitative methods in

28
00:02:28.050 --> 00:02:32.910
that they will allow a standard curve to produce definite endotoxin results.

29
00:02:33.900 --> 00:02:38.700
Now, the turbidimetric method has a measurement principle very similar to the gel

30
00:02:38.790 --> 00:02:43.740
clot method. In fact, our reagent is licensed to perform either method.

31
00:02:45.465 --> 00:02:51.360
But the turbidimetric method involves an absorbance reader

32
00:02:51.660 --> 00:02:55.020
measuring the onset of turbidity of that reagent.

33
00:02:55.410 --> 00:03:00.810
In contrast, we also manufacture chromogenic reagent in which a chromogenic

34
00:03:00.870 --> 00:03:06.570
substrate has been added to the reagent to measure the onset of this

35
00:03:06.601 --> 00:03:10.995
clotting enzyme more effectively using an absorbance reader.

36
00:03:11.700 --> 00:03:17.281
So these are the three general methods that the USP85 outlines in the bacterial

37
00:03:17.310 --> 00:03:21.451
endotoxin test, and we're going to be looking specifically at turbidimetric and

38
00:03:21.480 --> 00:03:22.800
chromogenic methods.

39
00:03:24.630 --> 00:03:30.570
Now, we're going to look at what the USP85 requirements of a quantitative method

40
00:03:30.585 --> 00:03:33.330
are, specifically of the reader itself.

41
00:03:33.451 --> 00:03:36.914
The reader needs to be a photometric incubating reader.

42
00:03:37.410 --> 00:03:42.675
It needs to be capable of measuring a negative control without activation.

43
00:03:42.690 --> 00:03:47.010
This may be stating the obvious, but of course, a negative control is supposed to

44
00:03:47.070 --> 00:03:51.840
be there to ensure that the environment is free of contamination.

45
00:03:51.930 --> 00:03:57.046
So only activation of the negative control should come about by

46
00:03:57.090 --> 00:04:02.460
the sample that we are attempting to measure, in which case would be

47
00:04:02.550 --> 00:04:08.101
endotoxin. Then whatever regression is used for the

48
00:04:09.330 --> 00:04:14.190
standard curve regression, the correlation coefficient must be greater than or

49
00:04:14.250 --> 00:04:19.710
equal to 0.980. Of course, if it's negative, it should be less than that.

50
00:04:20.010 --> 00:04:24.480
Basically, the absolute value of that correlation coefficient should be greater

51
00:04:24.540 --> 00:04:30.300
than 0.980. Samples must be tested with a positive product control

52
00:04:30.315 --> 00:04:35.310
spiked at or near the midpoint of the standard curve, and then measured positive

53
00:04:35.341 --> 00:04:41.220
product control results must lie within 50% to 200% of a true value.

54
00:04:41.281 --> 00:04:46.664
Now, many errors, whether that's technical, user, environmental,

55
00:04:47.250 --> 00:04:51.900
as well as product interference itself, could all come into the variability of the

56
00:04:51.960 --> 00:04:57.855
assay. So this 50% to 200% range, you can think of as the resolution of the test,

57
00:04:57.960 --> 00:05:03.720
but in a sense, it's that catch-all variability we allow for this biological

58
00:05:03.930 --> 00:05:04.531
assay.

59
00:05:06.570 --> 00:05:11.160
Now, all of our reagents are kinetic reagents.

60
00:05:11.370 --> 00:05:13.951
We measure them using kinetic methods.

61
00:05:14.400 --> 00:05:19.500
There are endpoint endotoxin reagents out there in which incubation takes place

62
00:05:19.560 --> 00:05:22.890
over a set amount of time, and then one measurement is taken.

63
00:05:23.460 --> 00:05:29.160
However, we use kinetic measurements for both our chromogenic and our

64
00:05:29.190 --> 00:05:35.160
turbidimetric techniques. Some of the advantages of this are there's no need

65
00:05:35.220 --> 00:05:40.951
to take the time yourself, at which point a plate would need a stop reagent

66
00:05:41.100 --> 00:05:47.010
added and then a measurement taken place, as well as a kinetic method allows for

67
00:05:47.100 --> 00:05:51.870
a wide standard curve range. And I visualize that here for you.

68
00:05:52.560 --> 00:05:58.201
First, you'll notice the negative control remains unreactive above our threshold

69
00:05:58.380 --> 00:06:03.660
OD, optical density. That's set at 0.015 for our reagent.

70
00:06:04.230 --> 00:06:08.730
Basically, that's the threshold at which an activation time is recorded.

71
00:06:09.360 --> 00:06:15.060
You can see I have four concentrations ranging from five to 0.05 EU per

72
00:06:15.210 --> 00:06:20.640
ml, that's just endotoxin units per milliliter, that have an activation time.

73
00:06:21.960 --> 00:06:27.826
So we have a measure of concentration that gives us a measure of activation time,

74
00:06:28.201 --> 00:06:32.670
and from those values, we can create a standard curve.

75
00:06:33.331 --> 00:06:38.701
Now, any regression is allowed by the Pharmacopeial standards, but our reagent

76
00:06:38.760 --> 00:06:42.510
specifically, we use a log-log linear reduction

77
00:06:43.710 --> 00:06:47.655
for the regression of that standard curve. So you can see that here.

78
00:06:47.670 --> 00:06:53.430
Basically, on a logarithmic graph, you see the CSE concentration on the

79
00:06:53.670 --> 00:06:59.341
x-axis, the onset time on the y-axis, and you can see the absolute value of

80
00:06:59.370 --> 00:07:03.405
linear regression that we received from that.

81
00:07:04.651 --> 00:07:09.690
That's an overview of the kinetic measurement, and then based off of unknown

82
00:07:09.720 --> 00:07:15.360
samples' activation times, we're able to plug those values into the standard

83
00:07:15.420 --> 00:07:19.680
curve to get measures of the endotoxin concentration.

84
00:07:22.230 --> 00:07:28.140
I now want to give you an overview of our kinetic endotoxin testing

85
00:07:28.230 --> 00:07:33.811
reagents that we provide. First of all, the PyroStar-ESF plate reagent.

86
00:07:34.081 --> 00:07:39.660
This is a kinetic turbidimetric reagent, and we validate it to

87
00:07:40.081 --> 00:07:45.930
produce for you a standard curve from 10 to 0.01 endotoxin units

88
00:07:46.020 --> 00:07:51.660
per milliliter. We also manufacture the Limulus Color-KY reagent,

89
00:07:52.110 --> 00:07:54.870
which is our kinetic chromogenic reagent.

90
00:07:55.500 --> 00:08:01.020
You can see one of the advantages here is that the quantitative range

91
00:08:01.201 --> 00:08:06.990
is from five all the way down to 0.0005 endotoxin units

92
00:08:07.050 --> 00:08:07.920
per milliliter.

93
00:08:09.151 --> 00:08:13.140
Now, both of these are Limulus amebocyte lysate reagent.

94
00:08:13.530 --> 00:08:19.081
Both of these fit the definition of an endotoxin testing reagent according to

95
00:08:19.140 --> 00:08:24.526
USP85. However, the United States Pharmacopeia has just voted to

96
00:08:24.630 --> 00:08:29.070
include a new chapter on bacterial endotoxin testing.

97
00:08:29.130 --> 00:08:34.740
That's chapter 86, and this chapter allows for synthetic reagents to be

98
00:08:34.831 --> 00:08:40.651
used. So I've also included on here our PyroStar-Neo reagent, which is a completely

99
00:08:40.681 --> 00:08:45.901
synthetic reagent, but it's measured by the kinetic chromogenic technique.

100
00:08:46.350 --> 00:08:50.460
In fact, all of the settings are applicable among those two reagents.

101
00:08:52.320 --> 00:08:58.260
It has a quantitative range from 50 to 0.001 endotoxin

102
00:08:58.320 --> 00:08:59.520
units per milliliter.

103
00:09:01.806 --> 00:09:07.507
Now, the USP, as we mentioned, has some requirements for quantitative

104
00:09:07.596 --> 00:09:13.027
methods. However, we as well as the reagent manufacturer have set additional

105
00:09:13.116 --> 00:09:18.336
requirements in our package inserts for accurate measurement with our reagent.

106
00:09:18.726 --> 00:09:24.696
So we validate that our reagent will provide for you results that are within

107
00:09:24.726 --> 00:09:30.576
specification if you meet the testing parameters that we require

108
00:09:30.606 --> 00:09:32.556
for valid plate reader measurements.

109
00:09:33.096 --> 00:09:37.926
So you can see those listed here for our kinetic chromogenic and kinetic

110
00:09:37.986 --> 00:09:43.716
turbidimetric reagents. Nearly identical requirements for each incubation

111
00:09:43.806 --> 00:09:47.496
temperature of 37 degree Celsius throughout the read.

112
00:09:47.616 --> 00:09:52.640
Of course, assay mode of kinetic. I mentioned that threshold optical density of

113
00:09:52.896 --> 00:09:54.306
0.015.

114
00:09:55.566 --> 00:10:01.206
We ask that you perform a shake at the beginning of the protocol

115
00:10:01.296 --> 00:10:06.666
before the reads start. The shake often we recommend around 10 seconds on a medium

116
00:10:06.816 --> 00:10:12.516
constant speed. Measurement wavelength is where these two methods do

117
00:10:13.926 --> 00:10:19.206
diverge from each other. Both require a measurement at 405

118
00:10:19.266 --> 00:10:24.996
nanometers. However, the chromogenic reagent also takes measurement at 650

119
00:10:25.086 --> 00:10:30.007
nanometers, and this signal is subtracted from the signal taken at the same time

120
00:10:30.066 --> 00:10:35.916
from 405 nanometers. This is to ensure that all of the noise from the general

121
00:10:35.976 --> 00:10:41.766
turbidity is subtracted and only the chromogenic portion

122
00:10:41.856 --> 00:10:44.646
of that signal is measured at that wavelength.

123
00:10:45.906 --> 00:10:51.126
We also recommend if your reader has a carriage speed setting, that that be set to

124
00:10:51.216 --> 00:10:55.956
slow. The measurement interval is 40 seconds for each of the reads.

125
00:10:56.016 --> 00:11:01.716
And then for the standard curve measurement criteria, that log-log linear reduction

126
00:11:01.731 --> 00:11:06.216
that I mentioned. So our package inserts outline these points.

127
00:11:06.277 --> 00:11:12.216
If you follow these points with a valid microplate reader, then we guarantee the

128
00:11:12.246 --> 00:11:13.926
results of the reagent.

129
00:11:15.936 --> 00:11:20.796
Now, what we have done is we've partnered with molecular devices to create a

130
00:11:20.856 --> 00:11:26.106
protocol for you that fits these criteria. So no need to set that up.

131
00:11:26.166 --> 00:11:29.286
It's automatically set up for you in the protocols.

132
00:11:29.616 --> 00:11:35.196
So what I want to do is look at some example data of a comparison of an

133
00:11:35.286 --> 00:11:40.716
assay run on two different readers, the molecular device reader, and a reader that

134
00:11:40.777 --> 00:11:45.996
we would use in-house that we currently use for the validation of the reagent to

135
00:11:46.086 --> 00:11:51.906
see the differences in the data, and the similarities, and also the

136
00:11:51.996 --> 00:11:56.527
validity of the report as it is generated.

137
00:12:00.096 --> 00:12:05.706
So first, I have data using our PyroStar® ESF reagent, which

138
00:12:05.856 --> 00:12:09.906
is that kinetic turbidimetric reagent.

139
00:12:10.326 --> 00:12:16.266
We ran the standard curve along the full range that we have validated the reagent

140
00:12:16.386 --> 00:12:21.007
at from 10 all the way down to 0.01 endotoxin units per milliliter.

141
00:12:21.996 --> 00:12:27.246
We've set up the reagent with the parameters that we would have

142
00:12:27.606 --> 00:12:33.126
specified in the protocols, and you can see the activation times and the linear

143
00:12:33.246 --> 00:12:37.596
regression equation that both readers would have produced.

144
00:12:38.406 --> 00:12:42.096
Now, we'll get to the measure of validity in a moment

145
00:12:43.716 --> 00:12:49.296
in the specifications that USP 85 requires of a valid standard curve.

146
00:12:49.356 --> 00:12:54.907
But I do want you to notice at the offset here, we are getting slightly different

147
00:12:55.056 --> 00:13:00.996
values for activation time. Now, you'll notice I haven't mentioned any specific

148
00:13:01.026 --> 00:13:06.231
requirement for activation time, and I'll explain the reason that's the case is

149
00:13:06.246 --> 00:13:11.526
because there are so many variables that can affect activation time, but are

150
00:13:11.616 --> 00:13:15.727
accounted for by the production of a standard curve.

151
00:13:15.757 --> 00:13:21.366
But we'll take note and I'll provide an explanation of why these differences are

152
00:13:21.426 --> 00:13:25.806
reflected, and you can see that slightly in the standard curve.

153
00:13:26.227 --> 00:13:30.996
So they're very similar values, but there is a slight lengthening of the numbers

154
00:13:31.056 --> 00:13:35.916
for the SpectraMax® ABS Plus microplate reader than this reader that

155
00:13:37.536 --> 00:13:41.886
we use in-house for the validation of the reagent.

156
00:13:43.776 --> 00:13:48.936
When we look at the data that was produced, we'll start to understand the slight

157
00:13:48.996 --> 00:13:54.126
difference in activation times. First of all, I want to mention the fact that the

158
00:13:54.186 --> 00:13:58.806
linear regressions for the two are very, very good, nearly identical.

159
00:13:58.926 --> 00:14:04.477
So we have a nearly identical correlation, and of course, over time, the averages,

160
00:14:04.566 --> 00:14:06.875
we'd see no difference between these.

161
00:14:08.316 --> 00:14:14.227
But one thing I do notice, and it may be hard to see on the screen, is that the

162
00:14:16.116 --> 00:14:21.666
SpectraMax® ABS Plus reader is showing a different value

163
00:14:21.727 --> 00:14:27.426
for the optical density of the final reaction that takes

164
00:14:27.456 --> 00:14:33.186
place. That's perfectly fine. Once again, we have no specified values for that.

165
00:14:33.216 --> 00:14:38.586
But what that does mean is that that measure of optical density is relative.

166
00:14:38.646 --> 00:14:40.926
It is unique to each of the readers.

167
00:14:40.956 --> 00:14:46.866
And so that reason right there with that threshold value of 0.015 is the reason we

168
00:14:46.926 --> 00:14:50.166
got slightly different activation times.

169
00:14:50.227 --> 00:14:55.596
However, we can, in a sense, cross-reference calculated values into the standard

170
00:14:55.626 --> 00:15:00.426
curve to eliminate some of these variables that would have come about by the

171
00:15:00.576 --> 00:15:03.157
optical density. And I show that right here.

172
00:15:04.146 --> 00:15:09.996
So basically what I have in the first graph is a measure of the true values

173
00:15:10.086 --> 00:15:15.276
in the terms of the control standard endotoxin concentrations that were used, and

174
00:15:15.336 --> 00:15:20.466
the values that were produced by the in-house reader and the SpectraMax® ABS Plus

175
00:15:20.586 --> 00:15:26.361
reader. So in a sense, basically, the raw activation times

176
00:15:26.407 --> 00:15:31.176
were used to create the standard curve, and then we plugged those activation times

177
00:15:31.296 --> 00:15:33.831
back into that standard curve that was produced.

178
00:15:34.535 --> 00:15:40.447
Now, it's just a fact of linear regression that the line of best fit is, in a

179
00:15:40.477 --> 00:15:45.996
sense, going to compromise through those data points to find that line of best

180
00:15:46.056 --> 00:15:51.486
fit. And so typically, the extreme ends will be, in a sense, biased in one

181
00:15:51.546 --> 00:15:56.046
direction, while the median points will be biased in another direction.

182
00:15:56.061 --> 00:16:01.356
And you see that here, where in the bottom graph, we're seeing difference from true

183
00:16:01.386 --> 00:16:05.796
value for the in-house reader and the SpectraMax® ABS Plus.

184
00:16:05.826 --> 00:16:09.876
We're seeing, in a sense, an under-reporting at the extremes and then an

185
00:16:09.966 --> 00:16:15.876
over-reporting at the means. That's just the fact that we have a linear regression

186
00:16:16.836 --> 00:16:20.046
calculation taking place here in that line.

187
00:16:20.106 --> 00:16:25.536
The data has a slight curve to it, and that line is approximating

188
00:16:26.197 --> 00:16:31.026
the regression for that. This is perfectly normal.

189
00:16:31.326 --> 00:16:36.276
Once again, we're seeing values well within the 50% to 200% range.

190
00:16:37.206 --> 00:16:42.846
That is the overall accuracy, or I should say, resolution of the

191
00:16:42.936 --> 00:16:48.157
assay. But what I want to focus on is what we're seeing here is difference from

192
00:16:48.186 --> 00:16:50.346
true value for these two different readers.

193
00:16:50.736 --> 00:16:56.197
Then a comparison of the difference between measurements, and in a sense, we take

194
00:16:56.227 --> 00:17:01.551
the values from the SpectraMax® ABS Plus reader and make a direct

195
00:17:01.596 --> 00:17:06.216
comparison to those endotoxin values from that in-house reader that we would use.

196
00:17:06.876 --> 00:17:11.181
You can see virtually no difference between those.

197
00:17:11.197 --> 00:17:15.757
Of course, there's variability as we will have in a physical world, but what we're

198
00:17:15.786 --> 00:17:19.236
seeing is basically exactly the same values.

199
00:17:19.716 --> 00:17:25.536
So what that's telling us is the readers are even picking up on the nuances of the

200
00:17:25.566 --> 00:17:30.937
endotoxin assay itself and are providing for us virtually identical results.

201
00:17:32.466 --> 00:17:37.866
The important thing, however, is that all test criteria are met for these

202
00:17:37.926 --> 00:17:42.876
results, and the difference between readers falls within that measurement error.

203
00:17:43.356 --> 00:17:48.426
Of course, the specified error of 50% to 200%, and then also the linear regression

204
00:17:48.546 --> 00:17:52.116
error that we were seeing for this specific standard curve.

205
00:17:54.756 --> 00:18:00.697
We also did a very similar comparison with the same readers, just with

206
00:18:00.756 --> 00:18:05.496
that chromogenic assay, and you can see here the standard curve range now

207
00:18:05.616 --> 00:18:10.956
encompasses the entire standard curve that we had, and you can see the activation

208
00:18:11.046 --> 00:18:16.086
times. Once again, we're seeing for the SpectraMax® ABS Plus reader a slight

209
00:18:16.206 --> 00:18:20.016
lengthening of those activation times.

210
00:18:20.706 --> 00:18:25.836
You can see the linear regression of the standard curve here that we see

211
00:18:27.697 --> 00:18:33.246
Here you can see the graphs of the time course and the activation that

212
00:18:33.276 --> 00:18:37.581
occurred. And you'll notice once again, we're seeing that difference in optical

213
00:18:37.716 --> 00:18:43.206
density, which is translating for us to that slight difference in activation times.

214
00:18:43.447 --> 00:18:49.296
Once again, no necessarily specified requirements for this, just something to

215
00:18:49.356 --> 00:18:53.796
observe and take note of. That's what we're seeing here because the readers are

216
00:18:53.826 --> 00:18:59.286
going to have different optical features, different ways of measurement.

217
00:18:59.466 --> 00:19:04.836
Optical density is a relative value that can vary based on the

218
00:19:04.896 --> 00:19:08.376
optics of the physical instruments themselves.

219
00:19:09.366 --> 00:19:14.435
But once again, when we plug in those true values, and

220
00:19:15.336 --> 00:19:19.056
we plug in the activation times against the standard curve to get the measured

221
00:19:19.086 --> 00:19:24.996
values, you can see here those differences that are produced, and you can see

222
00:19:25.056 --> 00:19:27.306
that linear regression error.

223
00:19:28.866 --> 00:19:33.697
The chromogenic reagent performs, in a sense, much better even than the

224
00:19:33.726 --> 00:19:39.606
turbidimetric reagent did. But even the difference between measurements lie within

225
00:19:39.666 --> 00:19:43.416
those as well. So once again, we see that

226
00:19:45.066 --> 00:19:48.426
the conclusion here is that all test criteria are met.

227
00:19:48.486 --> 00:19:53.706
We saw that linear regression, the correlation coefficient, absolute value above

228
00:19:53.916 --> 00:19:59.616
0.980, and then all measured values within that resolution of the test.

229
00:20:00.126 --> 00:20:05.437
And then even additionally, we see that the differences between the readers

230
00:20:06.426 --> 00:20:11.886
are within the error of the linear regression line itself when we plug the values

231
00:20:11.947 --> 00:20:13.146
back into each other.

232
00:20:14.676 --> 00:20:20.466
So in conclusion here, the Fujifilm Wako endotoxin test reagents and Molecular

233
00:20:20.526 --> 00:20:26.466
Device SpectraMax microplate readers with SoftMax Pro software provide a fully

234
00:20:26.586 --> 00:20:30.906
validated solution for your endotoxin testing needs.

235
00:20:31.956 --> 00:20:36.651
I want to say thank you for your time and allowing me to bring this presentation.

236
00:20:36.666 --> 00:20:42.246
I'm now going to turn the rest of the presentation over to Molecular Devices, and

237
00:20:42.276 --> 00:20:48.187
they're going to explain the details of the software protocol itself, especially

238
00:20:48.246 --> 00:20:49.986
focusing on the

239
00:20:51.187 --> 00:20:51.697
data

240
00:20:53.616 --> 00:20:58.356
compatibility in terms of data compliance with CFR and FDA

241
00:20:59.346 --> 00:21:03.096
GMP guidelines for that. Thank you so much for your

242
00:21:03.156 --> 00:21:05.886
time.

243
00:21:07.956 --> 00:21:11.676
I'm a senior application scientist at Molecular Devices.

244
00:21:11.736 --> 00:21:15.576
In my role with the company, I've worked to validate key assays on our microplate

245
00:21:15.606 --> 00:21:19.776
readers to meet the needs of scientists who are doing bacterial endotoxin testing

246
00:21:20.256 --> 00:21:24.621
and who need not only a plate reader that can generate the necessary data, but also

247
00:21:24.621 --> 00:21:28.236
analysis protocols that allow them to be compliant with regulations.

248
00:21:28.896 --> 00:21:32.496
I'm really happy to discuss with you today our collaboration with Fujifilm Wako

249
00:21:32.586 --> 00:21:36.306
Chemicals to provide solutions for endotoxin testing.

250
00:21:36.456 --> 00:21:37.866
Facts about our microplate readers.

251
00:21:38.346 --> 00:21:41.406
Molecular Devices has been building them for nearly 40 years.

252
00:21:41.825 --> 00:21:45.891
The picture on the top is the Emax, one of the OG plate readers.

253
00:21:45.937 --> 00:21:49.896
In the middle pic, you'll see a SpectraMax reader enjoying some lovely weather in

254
00:21:49.926 --> 00:21:54.606
Antarctica. And the bottom photo is the International Space Station, home to two of

255
00:21:54.621 --> 00:21:59.286
our SpectraMax M5 readers. So almost anywhere you go, you will find Molecular

256
00:21:59.346 --> 00:22:04.146
Devices plate readers. The earth where the reader I want to focus on is the

257
00:22:04.206 --> 00:22:09.350
SpectraMax ABS Plus. This single mode absorbance reader has some features that make

258
00:22:09.396 --> 00:22:11.676
it ideal for bacterial endotoxin testing.

259
00:22:12.471 --> 00:22:15.876
Monochromator-based wavelength selection means that you don't have to install

260
00:22:15.906 --> 00:22:20.226
different filters for each assay because you have wavelengths all the way from 190

261
00:22:20.286 --> 00:22:23.706
nanometer to 1,000 nanometer all included on board the reader.

262
00:22:24.516 --> 00:22:29.796
Temperature can be regulated from five degrees Celsius above ambient to 45 degrees.

263
00:22:30.306 --> 00:22:34.671
Of course, this range does take care of the endotoxin assays which are run at 37

264
00:22:34.671 --> 00:22:39.306
degrees. Kinetic reads are easy to set up, and as I'll show you, we have preset

265
00:22:39.336 --> 00:22:43.566
protocols that automate the analysis for Fujifilm's endotoxin assays.

266
00:22:44.256 --> 00:22:49.116
The ABS Plus also has eight channel optics, so eight wells are read simultaneously.

267
00:22:49.596 --> 00:22:53.796
This means that a 96 well microplate can be read in just about five seconds.

268
00:22:54.697 --> 00:22:58.386
And all of the SpectraMax readers come with SoftMax Pro software.

269
00:22:58.956 --> 00:23:03.065
As I mentioned, we provide pre-configured protocols that you can use with the

270
00:23:03.096 --> 00:23:06.020
software to analyze assay data automatically.

271
00:23:06.711 --> 00:23:11.046
The endotoxin protocols themselves, I would like to talk a little bit about SoftMax

272
00:23:11.076 --> 00:23:15.606
Pro software, which is our microplate reader control and data analysis software.

273
00:23:16.326 --> 00:23:20.301
Its protocol library has pre-configured protocols for all sorts of microplate

274
00:23:20.346 --> 00:23:26.316
reader assays, over 150 protocols in all, including for endotoxin testing, DNA

275
00:23:26.406 --> 00:23:30.216
quantitation, ELISA analysis, cell viability, et cetera.

276
00:23:30.666 --> 00:23:33.246
And they're all set up to simplify the data analysis.

277
00:23:34.026 --> 00:23:39.036
Compliance tools are available for regulated, which is GLP/GMP laboratories, and

278
00:23:39.096 --> 00:23:43.176
we've designed endotoxin protocols that can be used in either regulated or

279
00:23:43.206 --> 00:23:48.156
non-regulated spaces. This Molecular Devices software that provides compliance

280
00:23:48.216 --> 00:23:53.166
features for GLP and GMP practices, including several features that are really

281
00:23:53.197 --> 00:23:57.081
important for labs engaged in bacterial endotoxin testing.

282
00:23:57.126 --> 00:24:02.796
With this software, you can achieve full FDA 21 CFR Part 11 and Annex 11

283
00:24:02.826 --> 00:24:07.866
compliance. So for details on how, rather than discussing them here, I'll go ahead

284
00:24:07.896 --> 00:24:11.256
and refer you to a Lab Notes article that's on our website.

285
00:24:11.316 --> 00:24:15.697
And this lays out very clearly what Part 11 and Annex 11 are, what the

286
00:24:15.726 --> 00:24:20.616
responsibilities of regulated customers are, the impact of compliance versus

287
00:24:20.676 --> 00:24:24.816
non-compliance, and plenty of other useful information if you're interested.

288
00:24:25.356 --> 00:24:30.426
So some of the ways that SoftMax Pro GXC software supports compliance are with

289
00:24:30.456 --> 00:24:35.197
things like electronic records and electronic signatures that provide audit trails

290
00:24:35.226 --> 00:24:39.937
that can be accessed for review. You can also get automatically generated reports

291
00:24:39.996 --> 00:24:42.876
to fulfill various compliance requirements.

292
00:24:42.891 --> 00:24:47.270
Some data are also saved to a secure database, and the database can be backed up

293
00:24:47.286 --> 00:24:52.776
and restored. If you'd like more information on how SoftMax Pro GXC software can

294
00:24:52.836 --> 00:24:55.805
help you keep your lab compliant, please visit our website.

295
00:24:56.136 --> 00:25:00.786
There's a wealth of information, articles, white papers, tutorials, videos, et

296
00:25:00.846 --> 00:25:06.366
cetera. Now, part of the endotoxin testing solution that we provided along with

297
00:25:06.396 --> 00:25:10.266
Fujifilm, which is our endotoxin test protocols.

298
00:25:10.356 --> 00:25:14.106
These are among our newest protocol offerings, and as such, we do make them

299
00:25:14.166 --> 00:25:18.576
available on SpectraNet, which is our customer care portal on Molecular Devices

300
00:25:18.591 --> 00:25:22.716
website. It includes our knowledge base, which is a great place to get detailed

301
00:25:22.746 --> 00:25:25.701
information, articles, answers to many of your questions.

302
00:25:26.226 --> 00:25:31.116
SpectraNet is also where you'll find the SoftMax Pro protocol homepage, where you

303
00:25:31.146 --> 00:25:34.986
can find the latest versions of protocols that may not have been implemented yet in

304
00:25:35.016 --> 00:25:38.796
the SoftMax Pro software, but that we want to make available to anyone who would

305
00:25:38.826 --> 00:25:42.846
find them useful. This is where you'll find the two endotoxin protocols that we've

306
00:25:42.906 --> 00:25:47.391
co-designed with Fujifilm to work seamlessly with their kinetic turbidimetric and

307
00:25:47.406 --> 00:25:49.806
kinetic chromogenic endotoxin assays.

308
00:25:49.836 --> 00:25:54.021
Analysis that can be done with a SpectraMax ABS Plus reader for a kinetic

309
00:25:54.067 --> 00:25:59.406
turbidimetric endotoxin assay. It's a two-hour kinetic read with absorbance reads

310
00:25:59.466 --> 00:26:04.356
taken at 405 nanometers every 40 seconds for the duration of the read.

311
00:26:04.386 --> 00:26:08.886
This adds up to about 181 data points, giving us plenty of resolution to make

312
00:26:08.976 --> 00:26:11.286
accurate calculations from the data.

313
00:26:11.346 --> 00:26:16.326
The calculation or the reduction that is done on this kinetic data includes setting

314
00:26:16.386 --> 00:26:21.486
the first OD or optical density value to zero and then calculating the amount of

315
00:26:21.546 --> 00:26:25.986
time that's required for the subsequent OD readings to reach a specified threshold

316
00:26:26.346 --> 00:26:32.136
or onset OD. So the higher the endotoxin concentration is, the more rapidly this

317
00:26:32.196 --> 00:26:37.026
onset OD will be reached. Once we've generated all the data, a standard curve is

318
00:26:37.086 --> 00:26:42.457
plotted as mean onset time versus control standard endotoxin or CSE

319
00:26:42.516 --> 00:26:46.730
concentration. The curve fit that we've chosen here and that Timothy mentioned

320
00:26:46.776 --> 00:26:51.516
before is the log log fit. This is one of the many curve fits that are available in

321
00:26:51.546 --> 00:26:55.957
SoftMax Pro software. And for the standard curve in this assay, the correlation

322
00:26:55.986 --> 00:27:01.266
coefficient for R value for the standard curve must be at least 0.98.

323
00:27:01.866 --> 00:27:06.246
In this example from SpectraMax ABS Plus, you can see that the requirement is met

324
00:27:06.276 --> 00:27:07.476
and in fact exceeded.

325
00:27:13.236 --> 00:27:16.866
I'm going to walk you through the kinetic turbidimetric assay protocol.

326
00:27:17.421 --> 00:27:22.116
This one has been designed specifically for the Pyrostarr ES F plate assay from

327
00:27:22.177 --> 00:27:26.646
Fujifilm Wako Chemicals. And the first thing you'll see is the introduction, which

328
00:27:26.736 --> 00:27:31.416
outlines how the protocol is to be used and also gives information on how the assay

329
00:27:31.476 --> 00:27:36.042
is set up It also describes how to set up the protocol's template so that the

330
00:27:36.102 --> 00:27:40.813
samples in the different wells of the microplate will be analyzed appropriately.

331
00:27:40.872 --> 00:27:45.057
Instructions on entering information in the Reagent Info and Product Summary

332
00:27:45.192 --> 00:27:48.013
sections, which I'll show you in a moment, are also given here.

333
00:27:48.792 --> 00:27:53.022
The protocol are the User Entry and the Signature section.

334
00:27:53.112 --> 00:27:58.332
The User Entry section has areas for entry of things like the technician who's

335
00:27:58.362 --> 00:28:02.052
performing the assay, the date the assay was performed, and then various

336
00:28:02.112 --> 00:28:06.612
information on the reagents used, including the lot number for each and the

337
00:28:06.672 --> 00:28:08.052
expiration date for each.

338
00:28:09.643 --> 00:28:14.143
Product information is also entered, and that would include information such as the

339
00:28:14.203 --> 00:28:18.703
manufacturer, the product code, product description, product lot number, and the

340
00:28:18.763 --> 00:28:24.013
expiration date. There's also an area where assay-specific values like the

341
00:28:24.042 --> 00:28:28.393
endotoxin limit that you need to set for your test, the lowest concentration on the

342
00:28:28.453 --> 00:28:32.592
standard curve, and also the product dilution are entered in this slide.

343
00:28:33.643 --> 00:28:38.247
Now, the Signature section serves as an important record of tester information and

344
00:28:38.322 --> 00:28:41.982
test date, as well as approval and the date that it was approved.

345
00:28:42.492 --> 00:28:46.092
Securing these details is very important to regulatory compliance.

346
00:28:47.082 --> 00:28:51.792
See the Product Summary section, where statistical linear regression parameters are

347
00:28:51.806 --> 00:28:53.516
pulled directly from the standard curve.

348
00:28:53.682 --> 00:28:58.438
That's once the assay is run, and calculations are automatically applied so that

349
00:28:58.482 --> 00:29:00.897
you can see right away whether your results pass or fail.

350
00:29:01.692 --> 00:29:05.382
Over on the right-hand side, you'll see an overall judgment on the assay.

351
00:29:05.442 --> 00:29:11.112
Again, this is pass or fail. In this case, the product endotoxin concentration,

352
00:29:11.143 --> 00:29:16.453
which is 2.445, has exceeded the sample endotoxin limit of 0.5

353
00:29:17.143 --> 00:29:20.412
EU per mil, so it's clearly indicated as a fail.

354
00:29:21.192 --> 00:29:25.302
On the other hand, the positive product control spike recovery was within the

355
00:29:25.362 --> 00:29:29.127
acceptable range for the assay, so it's indicated as a pass.

356
00:29:29.772 --> 00:29:34.377
Other assay criteria, such as the absolute value of the correlation coefficient,

357
00:29:34.842 --> 00:29:39.492
also passed, but all parts of the assay must pass for the overall result to be a

358
00:29:39.552 --> 00:29:44.787
pass. If sections where the data are actually displayed and analyzed, starting with

359
00:29:44.802 --> 00:29:47.412
the plate or the Test Plate section.

360
00:29:47.442 --> 00:29:50.203
Here, you'll see the instrument settings used to run the assay.

361
00:29:50.667 --> 00:29:54.732
And in this section, you can also use the template to assign specific wells to

362
00:29:54.822 --> 00:30:00.522
groups such as standards, negative control, or positive product control.

363
00:30:00.582 --> 00:30:04.797
Once the plate's read, you'll see in the plate view all of the raw kinetic traces,

364
00:30:04.842 --> 00:30:08.893
and you can get a general overview of how your assay worked.

365
00:30:08.953 --> 00:30:13.497
At the bottom, the reduction settings are summarized, and on the right-hand side is

366
00:30:13.542 --> 00:30:18.117
the read information. That's where the parameters like instrument information is

367
00:30:18.132 --> 00:30:22.572
displayed, along with the read date, the time, the temperature, things like that.

368
00:30:22.992 --> 00:30:26.802
If for some reason you're finding that your results don't look as expected or

369
00:30:26.982 --> 00:30:31.272
questions arise later on, having a record like this could be very useful to

370
00:30:31.302 --> 00:30:35.232
compliance. My favorite section is the Standard Curve.

371
00:30:35.802 --> 00:30:40.122
Data for the standards are plotted here as mean onset time versus the control

372
00:30:40.212 --> 00:30:45.042
standard endotoxin concentration with a log log curve fit automatically applied to

373
00:30:45.072 --> 00:30:49.542
the graph. So curve fit results are displayed here, and you'll see some of the

374
00:30:49.602 --> 00:30:53.502
parameters that are also summarized in the Product Summary section that we looked

375
00:30:53.533 --> 00:30:58.513
at previously. The R squared value is also displayed here, and it lets you know how

376
00:30:58.572 --> 00:31:02.081
good, objectively, your standard curve linearity really is.

377
00:31:02.832 --> 00:31:07.332
Section will list the onset time for each well, along with the mean and the

378
00:31:07.393 --> 00:31:09.432
standard deviation and the percent CV.

379
00:31:10.242 --> 00:31:14.922
For each set of standard replicates, pass indicates a percent CV that's less than

380
00:31:15.102 --> 00:31:20.817
20. Below the table, you'll see a summary of statistical regression parameters, and

381
00:31:20.862 --> 00:31:24.882
there's a calculation to determine whether the absolute value of the correlation

382
00:31:24.972 --> 00:31:27.222
coefficient test passes or fails.

383
00:31:27.882 --> 00:31:31.513
Sections I'm showing here are for the negative control, the sample, and the

384
00:31:31.602 --> 00:31:37.152
positive product control. The negative control should confirm the absence of

385
00:31:37.212 --> 00:31:42.582
detectable endotoxin in the assay system, and as you can see here, it does.

386
00:31:43.572 --> 00:31:48.882
In the sample section, you get to see whether endotoxin that's detected is under

387
00:31:48.942 --> 00:31:50.652
the limit that you've set for your assay.

388
00:31:50.922 --> 00:31:55.272
In this example, although the assay replicates are good, you can see the percent CV

389
00:31:55.332 --> 00:31:58.122
is less than 20 for each set of replicates.

390
00:31:58.182 --> 00:32:02.712
The sample endotoxin concentration has in fact exceeded the test limit, so the

391
00:32:02.802 --> 00:32:08.712
overall result is a fail here. The positive product control, or PPC, is

392
00:32:08.772 --> 00:32:12.866
a test that's performed to make sure that the product you're testing doesn't have

393
00:32:12.912 --> 00:32:16.842
interfering factors in it that are going to skew the results of your test.

394
00:32:17.862 --> 00:32:21.703
So in this test, what you do is spike the test sample with a known quantity of

395
00:32:21.763 --> 00:32:26.592
endotoxin. As Timothy mentioned, this is typically a concentration of endotoxin

396
00:32:26.607 --> 00:32:30.987
that's near the middle of the standard curves range, and then you confirm that this

397
00:32:31.063 --> 00:32:36.492
is accurately quantified. So the test criteria here is that the calculated spike

398
00:32:36.522 --> 00:32:42.102
recovery needs to be between 50% and 200% of the actual concentration.

399
00:32:42.132 --> 00:32:46.632
In this example, the result is somewhat close, but it does pass.

400
00:32:46.752 --> 00:32:50.982
I really hope that Timothy and I have shown you how Fujifilm Wako Chemicals'

401
00:32:51.042 --> 00:32:56.082
endotoxin test reagents, in concert with Molecular Devices' SpectraMax microplate

402
00:32:56.143 --> 00:33:00.672
readers and the SoftMax Pro software that powers them, are a fully validated

403
00:33:00.703 --> 00:33:03.147
solution for your bacterial endotoxin testing.

404
00:33:03.777 --> 00:33:06.536
And I want to thank you very much for your time and your attention.