Professional Documents
Culture Documents
Hydrocarbon and Lipid Microbiology Protocols: Genetic, Genomic and System Analyses of Communities 1st Edition Terry J. Mcgenity
Hydrocarbon and Lipid Microbiology Protocols: Genetic, Genomic and System Analyses of Communities 1st Edition Terry J. Mcgenity
https://textbookfull.com/product/hydrocarbon-and-lipid-
microbiology-protocols-synthetic-and-systems-biology-
applications-1st-edition-terry-j-mcgenity/
https://textbookfull.com/product/hydrocarbon-and-lipid-
microbiology-protocols-synthetic-and-systems-biology-tools-1st-
edition-terry-j-mcgenity/
https://textbookfull.com/product/hydrocarbon-and-lipid-
microbiology-protocols-microbial-quantitation-community-
profiling-and-array-approaches-1st-edition-terry-j-mcgenity/
https://textbookfull.com/product/hydrocarbon-and-lipid-
microbiology-protocols-microbial-quantitation-community-
profiling-and-array-approaches-mcgenity/
Lipid Signaling Protocols 2nd Edition Mark Waugh (Eds.)
https://textbookfull.com/product/lipid-signaling-protocols-2nd-
edition-mark-waugh-eds/
https://textbookfull.com/product/biota-grow-2c-gather-2c-cook-
loucas/
https://textbookfull.com/product/the-genetic-manipulation-of-
staphylococci-methods-and-protocols-1st-edition-jeffrey-l-bose-
eds/
https://textbookfull.com/product/genomic-medicine-a-practical-
guide-laura-j-tafe/
https://textbookfull.com/product/the-ubiquitin-proteasome-system-
methods-and-protocols-thibault-mayor/
Terry J. McGenity
Kenneth N. Timmis
Balbina Nogales Editors
Hydrocarbon and
Lipid Microbiology
Protocols
Genetic, Genomic and System
Analyses of Communities
Springer Protocols Handbooks
Balbina Nogales
Department of Biology
University of the Balearic Islands
and Mediterranean Institute for Advanced
Studies (IMEDEA, UIB-CSIC)
Palma de Mallorca, Spain
All active cellular systems require water as the principal medium and solvent for their metabolic and
ecophysiological activities. Hydrophobic compounds and structures, which tend to exclude water,
although providing inter alia excellent sources of energy and a means of biological compartmental-
ization, present problems of cellular handling, poor bioavailability and, in some cases, toxicity.
Microbes both synthesize and exploit a vast range of hydrophobic organics, which includes biogenic
lipids, oils and volatile compounds, geochemically transformed organics of biological origin
(i.e. petroleum and other fossil hydrocarbons) and manufactured industrial organics. The underlying
interactions between microbes and hydrophobic compounds have major consequences not only for
the lifestyles of the microbes involved but also for biogeochemistry, climate change, environmental
pollution, human health and a range of biotechnological applications. The significance of this
“greasy microbiology” is reflected in both the scale and breadth of research on the various aspects
of the topic. Despite this, there was, as far as we know, no treatise available that covers the subject.
In an attempt to capture the essence of greasy microbiology, the Handbook of Hydrocarbon and
Lipid Microbiology (http://www.springer.com/life+sciences/microbiology/book/978-3-540-77584-
3) was published by Springer in 2010 (Timmis 2010). This five-volume handbook is, we believe,
unique and of considerable service to the community and its research endeavours, as evidenced by
the large number of chapter downloads. Volume 5 of the handbook, unlike volumes 1–4 which
summarize current knowledge on hydrocarbon microbiology, consists of a collection of experimen-
tal protocols and appendices pertinent to research on the topic.
A second edition of the handbook is now in preparation and a decision was taken to split off
the methods section and publish it separately as part of the Springer Protocols program (http://
www.springerprotocols.com/). The multi-volume work Hydrocarbon and Lipid Microbiology
Protocols, while rooted in Volume 5 of the Handbook, has evolved significantly, in terms of
range of topics, conceptual structure and protocol format. Research methods, as well as
instrumentation and strategic approaches to problems and analyses, are evolving at an unprec-
edented pace, which can be bewildering for newcomers to the field and to experienced
researchers desiring to take new approaches to problems. In attempting to be comprehensive
– a one-stop source of protocols for research in greasy microbiology – the protocol volumes
inevitably contain both subject-specific and more generic protocols, including sampling in the
field, chemical analyses, detection of specific functional groups of microorganisms and com-
munity composition, isolation and cultivation of such organisms, biochemical analyses and
activity measurements, ultrastructure and imaging methods, genetic and genomic analyses,
1
Adapted in part from the Preface to Handbook of Hydrocarbon and Lipid Microbiology.
v
vi Preface to Hydrocarbon and Lipid Microbiology Protocols
systems and synthetic biology tool usage, diverse applications, and the exploitation of bioin-
formatic, statistical and modelling tools. Thus, while the work is aimed at researchers working
on the microbiology of hydrocarbons, lipids and other hydrophobic organics, much of it will be
equally applicable to research in environmental microbiology and, indeed, microbiology in
general. This, we believe, is a significant strength of these volumes.
We are extremely grateful to the members of our Scientific Advisory Board, who have
made invaluable suggestions of topics and authors, as well as contributing protocols them-
selves, and to generous ad hoc advisors like Wei Huang, Manfred Auer and Lars Blank. We also
express our appreciation of Jutta Lindenborn of Springer who steered this work with profes-
sionalism, patience and good humour.
Reference
Timmis KN (ed) (2010) Handbook of hydrocarbon and lipid microbiology. Springer, Berlin, Heidelberg
Contents
vii
viii Contents
ix
x About the Editors
became paradigms of microbes that degrade organic compounds (Pseudomonas putida and Alcani-
vorax borkumensis). He has had the privilege and pleasure of working with and learning from some
of the most talented young scientists in environmental microbiology, a considerable number of
which are contributing authors to this series, and in particular Balbina and Terry. He is Fellow of the
Royal Society, Member of the EMBO, Recipient of the Erwin Schrödinger Prize, and Fellow of the
American Academy of Microbiology and the European Academy of Microbiology. He founded the
journals Environmental Microbiology, Environmental Microbiology Reports and Microbial Bio-
technology. Kenneth Timmis is currently Emeritus Professor in the Institute of Microbiology at the
Technical University of Braunschweig.
Abstract
Complex microbial ecosystems represent unique challenges to understanding, especially with regard to the
complex metabolic relationships that define the network of interactions that define the systems ecology of
an environment. Utilizing a wealth of available techniques we can now explore the genomic, transcriptomic,
proteomic, and metabolomic components of this system, with each component providing a window into a
stage of the network of interactions. Using these techniques we are just starting to map and validate the
mechanisms of catabolism, anabolism, and metabolite cross talk that enable microorganisms to sense and
interact with their environment. Translating this information into useful knowledge is the next major
challenge, with the end goal of producing models that can capture the variance and complexity of these
systems to faithfully reproduce observed characteristics of an ecosystem. We will explore some of these
techniques and tools in this section.
T.J. McGenity et al. (eds.), Hydrocarbon and Lipid Microbiology Protocols, Springer Protocols Handbooks, (2017) 1–4,
DOI 10.1007/8623_2014_5, © Springer-Verlag Berlin Heidelberg 2014, Published online: 19 November 2014
1
2 Jack A. Gilbert and Nicole M. Scott
References
1. Williams TJ, Cavicchioli R (2014) Marine meta- 2. Larsen PE, Gibbons SM, Gilbert JA (2012)
proteomics: deciphering the microbial metabolic Modeling microbial community structure and
food web. Trends Microbiol 22(5):248–260. functional diversity across time and space.
doi:10.1016/j.tim.2014.03.004 FEMS Microbiol Lett 332(2):91–98. doi:10.
1111/j.1574-6968.2012.02588.x
Genomic Analysis of Pure Cultures and Communities
Stepan V. Toshchakov*, Ilya V. Kublanov*, Enzo Messina,
Michail M. Yakimov, and Peter N. Golyshin
Abstract
Oil-degrading bacteria and their communities have been in focus of the research for the past few decades for
a number of reasons. First, this allows filling the voids in our knowledge on the major mechanisms
facilitating the oil biodegradation, to identify the key organisms playing significant roles in these processes
and, furthermore, to learn how to effectively manage their performance in situ to enhance the rates of
biodegradation. Historically, of a particular interest for genomics studies were the so-called marine hydro-
carbonoclastic bacteria, the petroleum biodegradation specialists with very restricted substrate profiles.
Apart from their utility in environmental cleanup, oil-degrading bacteria possess an array of enzymes and
pathways of a great potential for further biotechnological applications: biopolymers production, oxidation-
reduction reactions, chiral synthesis, biosurfactant production, etc. In this chapter we describe current
methods for genome and metagenome sequencing and annotation. Importantly, these are not limited to a
particular group of microorganisms and are thus almost universally applicable. We focused exclusively on
the methods and tools that everyone could use on a non-commercial basis. Due to the availability of
numerous alternative methods and approaches, we have arbitrarily chosen reliable protocols that can be
used by a common biologist without a great deal of computational biology background.
1 Introduction
T.J. McGenity et al. (eds.), Hydrocarbon and Lipid Microbiology Protocols, Springer Protocols Handbooks, (2017) 5–27,
DOI 10.1007/8623_2015_126, © Springer-Verlag Berlin Heidelberg 2015, Published online: 09 August 2015
5
6 Stepan V. Toshchakov et al.
1.1.2 Library Type There are two major kinds of DNA libraries generally used for de
novo (meta)-genomic assembly: fragment (shotgun or short-
insert) and mate-paired (or long-insert) library. Standard fragment
library protocol includes ultrasound or enzymatic fragmentation of
DNA molecule to short 150–800 nt long pieces, ligation of
sequencing adapters, size selection and PCR. Currently available
fragment library preparation kits are quite robust and allow prepa-
ration of the NGS library from DNA sample within 1 working day.
Final fragment library can be sequenced using single-end or
paired-end approach, depending upon the chosen platform.
Genomic Analysis of Pure Cultures and Communities 7
Table 1
Comparison of fragment and mate-paired library preparation protocols
1.1.3 Sequencing Each NGS platform currently available on the market can be
Platform characterized by several distinctive features, along with some
advantages and disadvantages. Thus, Roche 454 systems provided
longest reads among all second-generation platforms, SOLiD 5500
system gives a highest accuracy, Illumina HiSeq characterized
by deepest coverage, etc. All these characteristics of different
sequencing platforms are thoroughly analysed in a number of
excellent reviews [2, 3]. In this chapter, we would like the readers
to pay their attention mainly to tabletop sequencers, which are
considered as systems of choice for individual microbiology
research groups.
Introduced in 2011, both Ion Torrent Personal Genome
Machine (PGM) (Life Technologies) and MiSeq Desktop
Sequencer (Illumina) made next-generation sequencing technolo-
gies routine procedures in microbiology. While Illumina MiSeq
represents a reduced version of HiSeq possessing the same revers-
ible terminator sequencing chemistry, PGM is build on a new
principle based upon pH detection during incorporation of nucleo-
tides in growing chain. Current specifications of both systems are
summarized in Table 2. Despite MiSeq outperforms Ion Torrent by
throughput per run, short running time of PGM makes both
system comparable in terms of data production rate per day consid-
ering the 24/7 PGM load.
1.1.4 Minimal Coverage Reference-free de novo assembly requires significantly more cover-
age than reference-based (or mapping) assembly does. Generally
accepted value of read coverage for de novo assembly of pure
cultures is 100. This implies that for de novo assembly of bacteria
with genome of five million bases, one should generate at least
500 Mbp of sequencing data.
For the metagenomic assembly, one cannot make a reliable
judgment about total size of genomes presented in the sample.
In the literature, one can find some practical approaches based on
computational (statistical) methods typically counting selected
genes inside metagenome sequences in order to produce an
Table 2
Comparison of MiSeq and PGM sequencing systems (based upon February
2015 system specs)
2.1 General De novo assembly of genome sequence from a raw NGS data is a
Comments multistep process summarized in Fig. 1. Currently, one can find
numerous commercial and academic (usually, free) software
packages, each of which can execute particular assembly step. Aca-
demic software is usually designed to work with a specific kind of
data and for the specific application (Table 3). Below we describe
analysis pipeline used to determine genome sequence of pure
Quality
Raw files Quality control
trimming
Assembly
Scaffolding Gapfilling validation
Table 3
Commercial and freely available software packages for particular steps of de novo assembly pipeline
Galore! uk/projects/trim_galore/
Decontamination Deconseq http://deconseq.sourceforge.net/index. Any Both MacOS X/Linux [8]
html
Error correction Quake http://www.cbcb.umd.edu/software/ Illumina Standalone MacOS X/Linux [9]
quake/
Coral http://www.cs.helsinki.fi/u/lmsalmel/ 454/IonTorrent Standalone Linux [10]
coral/
De novo MIRA http://sourceforge.net/projects/mira- Any Standalone MacOS X/Linux [11]
assembly assembler/
Newbler http://www.454.com/products/analysis- Any Standalone Linux [12]
software/
Velvet https://www.ebi.ac.uk/~zerbino/velvet/ Any Standalone MacOS X/Linux [13]
Scaffolding SSPACE http://www.baseclear.com/genomics/ Any Standalone MacOS X/Linux [14]
bioinformatics/basetools/SSPACE
Velvet https://www.ebi.ac.uk/~zerbino/velvet/ Any Standalone MacOS X/Linux [13]
Filling of GapFiller http://www.baseclear.com/genomics/ Any Standalone MacOS X/Linux [15]
sequence gaps bioinformatics/basetools/gapfiller
GMcloser http://sourceforge.net/projects/gmcloser/ Any Standalone MacOS X/Linux Unpublished
Genomic Analysis of Pure Cultures and Communities 11
2.2 Raw Read Files In most cases, raw next-generation sequencing data are presented
as fastq files which store the reads as a nucleotide sequence with its
corresponding quality values for each nucleotide in a read (Fig. 2).
For sequencers using single nucleotide flow for sequence detec-
tion (IonTorrent and 454-based systems), data can also be pre-
sented as *.sff (Single Flow Format). If for some reason the user is
supplied just with *.sff files from the sequencing centre, it can easily
be converted to fastq with variety of tools (e.g. sff to fastq script
https://github.com/indraniel/sff2fastq)
2.3 Quality Control Extensive quality control of raw NGS data is a crucial step for all
kinds of downstream applications.
Despite there is a great variety of NGS quality control packages,
one of the easiest approaches would be to install and run FastQC
developed in Babraham Bioinformatics Institute (http://www.bioin
formatics.babraham.ac.uk/projects/fastqc/). FastQC is compiled in
Fig. 2 MiSeq raw data presented in fastq format. String containing read name (header) marked with (A);
nucleotide sequence (B); ‘+’ is a separator string; sequence of quality values in ASCII format (C )
12 Stepan V. Toshchakov et al.
Fig. 3 Results of FastQC analysis. In the example presented here, the read quality values are fine, but TruSeq
Adapter Index 5 sequences presented in the beginning of some reads result in the red flag on “Overrepre-
sented sequences”, “Per base sequence content” and “Kmer Content”
2.4 Quality Read To our experience, among numerous software packages used for
Trimming filtering and trimming next-generation sequencing reads, the pack-
age of choice should work with almost all kinds of sequencing data
and be highly adoptable in terms of filtering and trimming para-
meters. For routine NGS applications, we use Trim Galore! package
which allows to perform both quality and adapter trimming in
one step. It can be downloaded from Babraham Bioinformatics
website (http://www.bioinformatics.babraham.ac.uk/projects/
trim_galore/) and run by a command shell in Linux or MacOS X
system. In the first step, the software trims low-quality base calls
with thresholds specified by user. Next, the adapter sequences are
trimmed. By default, the Trim Galore! deletes sequences identical
to Illumina Universal adapter, but any other adapter sequences can
also be specified. It is very important to provide a level of stringency
Genomic Analysis of Pure Cultures and Communities 13
3.2 Non-coding RNA Quick rRNA-coding sequence predictions could be done using the
Predictions RNAmmer (http://www.cbs.dtu.dk/services/RNAmmer/ [27]).
RNASpace server (http://rnaspace.org/ [28]) and Rfam (http://
rfam.xfam.org/ [29]) database and search tool (similar by architec-
ture to Pfam, see below) are more convenient for more compre-
hensive search and annotation of non-coding RNAs. The fast way
of tRNA sequence prediction is to use the tRNAscan-SE web-based
tool [30, 31].
4.1 General Currently, plenty of tools are available, both online and offline.
Comments Actually, each step could be accomplished by different tools with
similar functionality. Here, we introduce some of them, which are
the most convenient to our experience. While the majority of tools
are located individually, there are several multifunctional servers for
comprehensive large-scale annotation and analysis, like IMG/ER
and IMG/MER or RAST and MG-RAST for genomic and meta-
genomic annotation, respectively. While the growing functionality
of the servers makes it possible to perform the majority of analyses
on the remote servers, it is still necessary to use different individual
applications in frames of further vigorous manual curation steps.
From our experience, the most prominent strategy is to use the
servers for structural annotation and initial steps of automatic
functional analysis using the results for following manual curation
Genomic Analysis of Pure Cultures and Communities 17
4.3 Genomic/ The ORF predictions could be performed using the above servers.
Assembly Structural In the RAST, there are two pipelines available: RAST ORF predic-
Annotation tion and GLIMMER ORF prediction. RAST has more options, but
usually has a higher overprediction rate; thus, we recommend
GLIMMER.
4.4 Analysis of De When ORFs were predicted, several approaches could be used for
Novo Annotated further protein function verification. In general, these approaches
Genomes/ are individual protein prediction, metabolic pathway predictions
Metagenomes and gene colocalization and coregulation. The latter is an impor-
tant part of analysis; however, we are not going to discuss it since it
is well reviewed in several papers [34–36]. For this, ORFs must be
translated into polypeptides due to genetic code degeneracy.
4.4.1 Basic Local Two strategies are (a) to BLAST entire database with your protein
Alignment Search Tool query and (b) to BLAST your genome/metagenome with a char-
(BLAST) [37] acterized protein.
For the first one, there are many BLAST servers available, both
online and offline, the most popular being at NCBI/NIH (http://
blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM¼blastp&PAGE_
TYPE¼BlastSearch&LINK_LOC¼blasthome) and Uniprot
(http://www.uniprot.org/blast/) BLAST servers. In most cases,
the default parameters are suitable. It is important to check the e-
values and to define acceptable the cut-offs, e.g. (below 0.01 or
below 106) of the hits and the coverage of both hits and query.
Ideally, for full-length single-domain protein search, the coverage
should not be far from 100%. The main “need-to-know” features
are:
(a) GenBank server supports multiple-sequence query, while Uni-
prot does not support this.
(b) Genbank server has advanced algorithms of analysis, e.g.
position-specific iterated PSI-BLAST [38], pattern-hit
initiated PHI-BLAST [39] and domain enhanced lookup
time accelerated DELTA-BLAST [40] which are suitable for
proteins, characterized by a low similarity to those present in
the databases.
Genomic Analysis of Pure Cultures and Communities 19
4.4.2 HMM and Pfam Another approach for protein functional characterization is a search
for conserved domain families against the curated Pfam family
database, constructed based on multiple-sequence alignment and
hidden Markov models (HMMs). It should be mentioned that
several servers as dbCAN (see above) or HMMER (see below)
provide both BLAST and HMM search tools.
Another random document with
no related content on Scribd:
"O sir, you wouldn't be so wicked, surely!" Dick broke in, in
accents of alarm. "We should starve outright, I believe,
without mother's Wednesday and Saturday earnings at the
Manor House. And the children ain't half fed as it is!" He
wound up with another flood of tears.
"Then hold your tongue, now that you know what your
silence is worth," replied Stephen. "I'm sure you needn't
make such a cry-baby of yourself. I haven't hurt you, and
I've given you a jolly little box."
"But the box isn't any use to me," Dick argued. "Please—
please give me back my shilling, Master Stephen. 'Tis
dreadful to be hungry; and mother started off to work this
morning without a bit of anything inside her lips, because
she knew if she ate breakfast there wouldn't have been
enough for the little ones."
"O Dick!" the little girl exclaimed, "What a long time you've
been! And how red the wind has made your poor eyes look
—just as if you had been crying!"
CHAPTER IV.
TEN SHILLINGS REWARD.
"Indeed I do!" sighed Mrs. Wilkins, and a hot tear fell upon
her work; she was knitting to-night by the uncertain light of
the fire. "Life's a struggle at the best of-times for poor
people," she went on; "and when the father of a family is
taken, it's bound to go hard with those he leaves behind."
"I don't need the use of my eyes to knit, dear," was the
widow's return. "If I was sewing, 'twould be different."
"Then you really think we shall have to part with him?" cried
Dick. "Oh! God must be very cruel if He lets it come to that.
I know our Stranger wouldn't ever love other folks as he
loves you and me and the children. And if we sold him or
gave him away, his new owner might kick him about as—as
some people do their dogs."
"Well, there's all next month for us to look around and try
to serape the money together, dear," the widow summoned
heart enough to remind her little boy. "As long as it's paid
by the last day of January it will be in time; and if 'tis right
for us to keep our dog, why then we shall find ways and
means for doing so. Don't fret, child, more than you can
help. Whatever happens will be sure to be for the best.
Now, dry your eyes, and we'll have our supper cosy like in
front of the fire. If you lose heart, Dick, what'll become of
us all?"
But though the old year died and the new one took its
place, no sign of better fortune could Mrs. Wilkins or Dick
see. Stranger must be disposed of—that seemed certain
beyond a doubt; and if no one could be induced to offer him
a home, why then he would have to be killed. It would be
terrible indeed to part with so faithful a friend.
"Dick, Dick, my little boy, what's the matter with you? Are
you ill?" demanded Mrs. Wilkins; for the small face at her
side had grown suddenly as pale as death, and the child
had clutched convulsively at her arm.
"What's true, Dick? I don't know what 'tis you're talking of."
"I didn't tell you because I didn't find it, and I could not
bring myself to worry you by saying how it got into my
hands," was the child's admission. Then, as they walked on
side by side in the direction of Leigh Grange, Dick narrated
the story of his meeting in the woods with Stephen Filmer,
adding, "And I thought God was so cruel to let that great
bully rob me of the shilling when I wanted it so badly. I little
dreamed that things would turn out as they have."
And now that the silver lining had appeared to his cloud,
Dick laughed merrily at the thought of how vexed the
squire's son would be when he discovered what he had lost
by not being able to restore the box.
CHAPTER V.
DICK'S INTERVIEW WITH THE COLONEL.
"Please, sir, I've brought back the match-box that you lost
some weeks ago," said Dick Wilkins, his heart beating so
loud that he fancied his questioner must hear it.
"O sir! O sir! Please to believe me when I tell you all about
it," sobbed poor Dick, "'cos Master Stephen's treated me
shameful, he has! He's the biggest bully in the place, and
he stole a shilling from me when he found me alone in the
woods."
"I shall pay the tax for our Stranger, sir. We should have
had to get rid of him if—if it hadn't been for this."
"So he's stayed with you ever since! I believe I have seen
him in the village on several occasions—a handsome
creature he seemed too."
"Wait, Stephen," said he; "I wish to speak to you about the
lost match-box that you assured me you would bring back
to-morrow. This lad has already brought it to me. What light
can you throw upon the matter?"
"I found the match-box," answered the boy sulkily. "I only
lent it to Dick Wilkins, and I suppose he's been dishonest
enough to claim the reward."
"Oh!" cried Dick, in shocked accents. "Oh! how can you say
so, Master Stephen?"
And the squire's son passed out of sight, for once in his life
really frightened and abashed.
"O sir!" gasped Dick, when he had gone, "I'm 'fraid he'll do
some mischief even now if he can. Supposing he should get
my mother out of her work at the Manor House! We should
starve! And—and our landlord's a hard man, he is!"
"Have you seen him, sir?" asked little Dick, scarcely able to
suppress his anxiety.
CHAPTER VI.
HARD TIMES.
How proud Dick felt next day when he walked into the
grocery establishment that was also the post-office, laid his
half-sovereign on the counter, and said he had come to pay
his dog-tax. Stranger was with him, and in such high spirits
that he found it hard to believe the dog did not understand
the nature of their errand.
"So you're not going to get rid of the retriever after all,
then," remarked the post-mistress, after filling in and
handing Dick the receipt for his money.
"No," said the little boy; and then he pointed at the notice
that had not yet been removed from the window, and
added, "That's how I got my half-sovereign, Mrs.
Mortimore. The colonel gave it to me for bringing his match-
box back to him last evening."
"Thank you very much for the tea," said Dick gratefully.
"Mother'll be glad of it, I'm sure." And with this he turned
towards the entrance of the shop, and would have gone his
way had not the talkative post-mistress called him back to
the counter again.
"If you take my advice, Dick Wilkins," she went on, "you'll
get that mother of yours to go and see the doctor. She's a
failing woman—you mark my words. Get Dr. Rogers to give
your mother something—there's a good boy!—or, in my
belief, you won't have a mother to care for you much
longer."
"Do you feel bad this evening?" asked the boy in anxious
tones. "I mean—does your side ache worse than usual?"
"No, dear, not worse than usual. Why, Dick, folks would
think I was a grand body, if they knew how careful you were
of me."
"I want you to see the doctor, mother. You do look ill and
bad!" declared Dick gravely.
"That's nonsense! It's the cold that nips me up," was the
prompt return. "'Tis freezing hard to-night again. I shouldn't
be surprised if the ice on the lake bears soon. Then you and
the children'll be able to go and watch the skating between
whiles. Lord Bentford is certain to throw his grounds open
to the public as usual. O—oh!"
"What made you cry out like that? Why, you've got your
hand tied up! What's amiss with it?"
"There's a sore place on one of the fingers; and when I
knocked it against the table, it made me cry out. 'Twill be
easier in a minute;" and Mrs. Wilkins turned her face aside
that Dick might not see it was drawn with pain.
"How long has your finger been bad?" the little boy
demanded.
"And you've been working with it sore all day!" cried Dick,
in much concern. "Hasn't it pained you, mother?"
Then all at once a bright idea flashed into Dick's mind. To-
morrow would be Saturday, and school holiday. He would
put a gimlet in his pocket, go to Lord Bentford's lake, which
by now was bearing, and try to earn a few coppers by
putting on the gentlefolks' skates. He would not breathe
one word of his intention to any one; no, not even to his
mother. So he went supperless to bed that night, full of
hope for the success of his new venture on the morrow.
CHAPTER VII.
A GALLANT RESCUE.
Not till this minute had it dawned on him that his dog had
followed. Had he loitered on his way, or even glanced once
behind, he must certainly have seen Stranger stealthily
tracking him. But he had done neither; and now as he
stared in vexation at the animal, his commonsense told him
that he must take him home before Lord Bentford or his
gamekeepers had a chance of seeing him.
Dick trembled at the threat, and set off after his wayward
property. But the ice was slippery, and he fell once or twice
and hurt himself badly. He had just picked himself up, when
a piercing shriek rang through the air, and was followed by
a woman's scream of alarm and a man's loud shout for
help.
"A rope, a rope!" some one was calling. "Bring a rope this
minute. There's a child in the water, near the boathouse,
where the ice has been broken for the swans. Quickly,
quickly, or we shall be too late!"
CHAPTER VIII.
STRANGER'S MISSION FULFILLED.