3841Circulating Nucleic Acids as Liquid Biopsies and Noninvasive Disease Biomarkers CHAPTER 490

p = 0.016

0.0

0.2

0.4

0.6

0.8

ACR grade ≥2

ACR grade 0

ACR grade 1

%ddcfDNA

p <0.001

0

1

2

3

AMR ≥2

AMR0

AMR1

%ddcfDNA

A B

p <0.001

0.0

0.2

0.6

0.8

None Moderate Severe

%ddcfDNA 0.4

Mild

C Severity of LVEF drop

FIGURE 490-5 Correlation of percentage of donor-derived cell-free DNA (%dd-cfDNA) measures with

heart transplant biopsy-graded rejection and allograft dysfunction by echocardiography. A. %dd-cfDNA

in relation to severity of acute cellular rejection (ACR) by histopathologic interpretation of endomyocardial

biopsy. Grade 0 rejection includes both ACR grade 0 and antibody-mediated rejection (AMR) grade 0.

P value obtained by generalized estimating equation approach comparing all categories. B. %dd-cfDNA in

relation to severity of AMR by histopathologic interpretation of the endomyocardial biopsy. Grade 0 rejection

includes both ACR grades 0 or 1 and AMR grade 0. P value obtained by generalized estimating equation

approach comparing all categories. C. %dd-cfDNA in relation to severity of allograft dysfunction measured

by echocardiography. Allograft dysfunction was defined as a reduction of left ventricular ejection fraction

(LVEF) by ≥5% and was further stratified by severity based on the magnitude of LVEF decline as no (<5%),

mild (5% to <10%), moderate (≥10% to <15%), or severe (≥15%) allograft dysfunction. P value obtained by

generalized estimating equation approach comparing all categories. (Reproduced with permission from

S Agbor-Enoh et al: Cell-free DNA to detect heart allograft acute rejection. Circulation 143:1184, 2021.)

Finally, the investigators demonstrated higher dd-cfDNA levels

associated with antibody-mediated rejection compared to cellular

(T-cell–mediated) rejection, as well as different cfDNA fragment

types and lengths during these two rejection processes, suggesting

that dd-cfDNA patterns could potentially help differentiate these two

rejection subtypes.

Parallel studies in lung, kidney, and liver transplantation have also

confirmed the utility of dd-cfDNA as a noninvasive acute rejection

biomarker.

A large, multicenter, prospective study in kidney transplantation (Diagnosing Acute Rejection in Kidney Transplant Recipients

[DART]) enrolled 384 patients, on whom 107 biopsies of the transplanted kidney were performed. The dd-cfDNA level discriminated

between biopsies showing any acute rejection and controls (no histologic rejection), with an AUC of 0.74, sensitivity of 59%, and specificity of 85% for acute rejection at a dd-cfDNA cutoff of ≥1.0%. Since

publication of the DART study, other investigations

of dd-cfDNA in kidney transplantation have shown

similar results, with notably high dd-cfDNA levels (median 1.4–2.9%) in patients with antibodymediated rejection and poorer diagnostic accuracy

for patients with early (Banff 1A or borderline)

T-cell–mediated rejection. Studies have also shown

elevated dd-cfDNA levels in patients who develop

donor-specific antibodies, even in the absence

of antibody-mediated rejection. In kidney transplantation, dd-cfDNA could be elevated due to

pathologies other than rejection (e.g., BK virus

nephropathy and pyelonephritis) and is more likely

a marker of severe graft injury rather than rejection

alone. Thus, follow-up testing to determine the

cause and nature of graft injury or use of dd-cfDNA

surveillance primarily in patients at high risk of

allograft rejection would be reasonable.

Studies in liver transplantation have consistently

demonstrated elevated dd-cfDNA levels during

episodes of biopsy-proven acute rejection. As in

kidney transplantation, levels appear to rise before

development of clinical manifestations of acute

rejection, with data demonstrating elevated levels

4–6 days before a rise in transaminase enzyme

levels and 8–15 days before biopsy confirmation of

rejection. It is important to note that steady-state

dd-cfDNA levels vary between transplanted organs,

likely relative to the organ’s mass, cellularity, or

vascularization, with levels of 5–10% seen in stable

liver transplant recipients and levels <0.1% seen in

heart transplant patients.

Similar patterns of dd-cfDNA have also been

seen after lung transplantation, with elevated levels

detected weeks prior to and at the time of transbronchial biopsy demonstrating acute rejection.

As in heart and kidney transplantation, dd-cfDNA

levels correlate with severity of the rejection event,

with greater discriminatory power seen for severe

rejection and higher levels seen in antibodymediated rejection compared to cellular (T-cell–

mediated) rejection.

Donor-derived cfDNA levels also appear to have

prognostic value, even in the absence of acute

allograft injury. In a lung transplant cohort, patients

with the highest average dd-cfDNA levels within

the first 3 months after transplant, in the absence of

rejection, were at high risk for subsequent chronic

allograft dysfunction and allograft failure. Similarly,

in kidney transplantation, donor-derived cfDNA

levels at hospital discharge were shown to predict

graft dysfunction at 1 year after transplant. As such,

average dd-cfDNA levels over time may reflect

cumulative graft injury and related long-term adverse outcomes.

Over the past few years, targeted next-generation sequencing assays

(AlloSure [CareDx, Inc.] and Prospera [Natera]) have been developed

to quantify dd-cfDNA levels after transplantation. These assays are

composed of panels of highly polymorphic SNPs that, when sequenced

and quantified, enable differentiation of donor and recipient cfDNA

molecules using bioinformatics approaches. Such assays do not require

donor or recipient genotyping, are rapid and relatively cost effective

(particularly when compared to whole genome sequencing), and are

currently available for clinical use.

In summary, dd-cfDNA can be reliably detected in the plasma of

transplant recipients. Levels of dd-cfDNA are high immediately after

transplantation but rapidly fall to a baseline level within 2 weeks, after

ischemia-reperfusion injury has subsided. These dd-cfDNA levels

then remain at stable low levels in the absence of graft injury, with

baseline levels varying by organ type, likely related to the mass of the

3842 PART 20 Frontiers

transplanted organ. Most studies performed to date have demonstrated

rising levels with development of acute rejection and other forms

of allograft injury, with highest discriminatory power for antibodymediated rejection and severe grades of acute cellular rejection, and

have shown that levels rapidly fall after successful rejection treatment.

dd-cfDNA testing cannot completely replace biopsies of the transplanted organ, which provide valuable immunohistochemical and

molecular diagnostic information that helps guide clinical treatment.

Rather, this noninvasive assay reduces the need for surveillance biopsies

and enables clinicians to selectively perform biopsies on patients for

whom they are truly indicated, with attendant improvements in safety,

cost, and patient satisfaction. The ultimate goal is to use dd-cfDNA to

“personalize” immunosuppressive management by enabling weaning of

therapy in stable patients with low dd-cfDNA levels and augmentation

of treatment in patients with rising levels suggestive of early graft injury.

■ MONITORING MICROBIAL DIVERSITY AND

INFECTION AFTER TRANSPLANTATION

Immunosuppressive therapies reduce the risk of allograft rejection

but increase the transplant recipient’s susceptibility to infectious complications. The balance between level of immune system suppression

and the competing risks of rejection and infection is delicate, and the

desirable “therapeutic window” for patient treatment is narrow. Furthermore, diagnosis of infectious complications after transplantation

is challenging, as patients are susceptible to a wide array of pathogens

and common diagnostic tests rely on an a priori hypothesis for the

source of infection and performance of specific antigen-based or PCRbased molecular tests. Hypothesis-free sequencing of the microbial

component of transplant recipients’ cfDNA offers a window into the

state of the immune response and potential diagnosis of infectious

complications.

In 2013, De Vlaminck et al performed shotgun sequencing of heart

and lung transplant recipient cfDNA followed by alignment to reference human and microbial databases. This work showed that 2% of

cfDNA sequences were microbial in origin, including viral, bacterial,

and fungal genomes. They then studied the microbiome composition

in plasma at different levels of taxonomic classification and showed

that viruses (73%) are more abundant than bacteria (25%) and fungi

(2%). Subsequent testing showed quantitative agreement between viral

counts as measured by sequencing and quantitative PCR assays for

common posttransplant viral pathogens such as herpesviruses.

Available data on clinical drug treatment and dosage, including calcineurin inhibitor use and antiviral prophylaxis, were used to analyze

drug-microbiome interactions. The investigators found that the structure of the virome is exquisitely sensitive to drug use and dosage—high

doses or levels of immunosuppressive drugs give rise to a virome

dominated by Anelloviridae or torque teno viruses, which are ubiquitous viruses that have no known pathogenic role in humans but have

been shown to replicate in the setting of immune system suppression.

Similarly, the relative amount of herpesviruses declines substantially

after introduction of the prophylactic antiviral agent valganciclovir

(Fig. 490-6). Using this approach, the investigators demonstrated dramatic changes in the viral composition of the microbiome in response

to introduction and weaning of immunosuppressive and antiviral therapies after transplantation.

Using a similar approach, De Vlaminck and colleagues analyzed

nonhuman cfDNA as a hypothesis-free approach to test for infectious

complications after lung transplantation. They showed that levels of

cytomegalovirus-derived sequences in cfDNA correlated well with

clinical testing results (AUC 0.91). They also identified adenovirus,

polyomavirus, herpesvirus, and microsporidia sequences in patients

with clinical infections that had eluded diagnosis, suggesting that

10%

3%

1%

60%

39% 31% 21%

30%

15%

34%

0–3

0–300 300–600

Antiviral drugs (Valganciclovir) (mg)

Immunosuppressant (Tacrolimus) ng/mL

600–900

3–6

6–9

9–12

12–15

72%

84%

14%

12%

54%

18%

80%

94%

2%

82%

4%

2%

87%

2%

82%

6%

75%

13%

87%

86%

Herpesvirales

Caudovirales

Adenoviridae

Anelloviridae

Polyomaviridae

Poxviridae

Retroviridae

Other

FIGURE 490-6 Relative viral genomic abundance as a function of drug dose in a cohort of heart and lung transplant recipients. Mean virome composition for patients

treated with the immunosuppressant tacrolimus (47 patients, 380 samples) as a function of antiviral drug dose (valganciclovir) and concentration of tacrolimus measured in

blood. To account for the delayed effect of the virome composition on drug dose, the data on drug doses were window average filtered (window size, 45 days). Herpesvirales

and caudovirales dominate the virome when patients receive low doses of immunosuppressants and antiviral drugs. Conversely, anelloviridae dominate the virome when

patients receive high doses of these drugs. (Reproduced with permission from I De Vlaminck et al: Temporal response of the human virome to immunosuppression and

antiviral therapy. Cell 155:1178, 2013.)

3843Circulating Nucleic Acids as Liquid Biopsies and Noninvasive Disease Biomarkers CHAPTER 490

Time period post transplant (months)

0

0

50

100

3 6 9 12 14 >15

Abundance (%)

Herpesvirales

Caudovirales

Adenoviridae

Anelloviridae

Polyomaviridae

Poxviridae

Retroviridae

Other

FIGURE 490-7 Viral genome abundance at the family and order level of taxonomic classification

for different time periods after heart and lung transplantation. The fraction of anelloviridae

expands rapidly in the first several months after transplantation. The fractions of herpesvirales,

caudovirales, and adenoviridae decrease in that same time period. After 6 months, the opposite

trends are observed. (Reproduced with permission from I De Vlaminck et al: Temporal response of

the human virome to immunosuppression and antiviral therapy. Cell 155:1178, 2013.)

No ACR Mild ACR Moderate

ACR

10

p = 0.003 p = 0.005

2

No CMV

A B

CMV

Infection

CMV

Disease

4

6

TTV DNA (log10 copies/mL)

8

10

2

4

6

TTV DNA (log10 copies/mL)

8

FIGURE 490-8 TTV (torque teno virus, anellovirus) load during (A) episodes of acute cellular rejection

(ACR) and (B) cytomegalovirus (CMV) infection and disease after liver transplantation. (Reproduced with

permission from P Ruiz et al: Torque teno virus is associated with the state of immune suppression early after

liver transplantation. Liver Transpl 25:302, 2019.)

cfDNA sequencing can be a powerful approach for nonbiased diagnosis of opportunistic infections after transplantation. However, several

obstacles to clinical implementation must first be overcome, including

the establishment of pathogen-specific thresholds to discriminate

between colonization, infection, and disease.

■ MONITORING HOST IMMUNITY

As mentioned previously, De Vlamink and colleagues demonstrated

that the Anelloviridae family of viruses predominated in the virome

of lung transplant recipients, as analyzed by plasma cfDNA sequencing, and accounted for 68% of the total viral population. The relative abundance of anellovirus cfDNA is exquisitely sensitive to

immunosuppression, ranging from <5% prior to transplant (e.g., in

the nonimmunosuppressed state) to 84% during

months 4.5–6 after transplantation, when immunosuppression is most profound. Starting at 6

months after transplantation, the relative amount

of anellovirus cfDNA declines, corresponding to

a reduction in intensity of immunosuppression

(Fig. 490-7).

The observation that anellovirus cfDNA levels rise and fall with degree of immune system

suppression gave rise to the hypothesis that the

abundance of this ubiquitous family of viruses

could be used as a marker of the overall strength

of the immune response. Indeed, a global marker

of the net state of immunosuppression has eluded

transplant physicians to date, who currently quantify doses and levels of individual immunosuppressive drugs but cannot reliably assess for their

effect or potency in combination. In support of

this hypothesis was the subsequent observation

that anellovirus load was lower (reflecting less

immunosuppression) in lung transplant recipients

who developed acute rejection and was relatively

high (indicating more profound immunosuppression) in patients who had a rejection-free

posttransplant course, suggesting that the anellovirus cfDNA load could be a surrogate marker of

immunocompetence.

Subsequent studies have confirmed these

observations after different solid organ transplant

procedures (Fig. 490-8). A prospective lung transplant

study showed a hazard ratio (HR) of 5.05 for infectious

complications and an HR of 0.48 for acute rejection with

every log10 increase in anellovirus load. Similar results

have been shown after liver and kidney transplantation,

with higher anellovirus levels associated with opportunistic infections and malignancy (sequelae of chronic

immunosuppression) and lower levels associated with

acute rejection and chronic allograft dysfunction. While

much work remains to be done in this arena, including

standardization of anellovirus assays and quantification,

prospective studies with well-defined sampling schedules

and endpoints, and improved understanding of genotype-specific effects on host immunity, the potential to

use anellovirus cfDNA load as a marker to titrate and

personalize immunosuppressive therapy may be within

reach.

In summary, cfDNA sequencing has myriad clinical

applications after solid organ transplantation. Donorderived cfDNA assays are currently available for acute

rejection monitoring and are being used for noninvasive

surveillance after heart, lung, and kidney transplantation.

These assays are highly sensitive, provide early detection of allograft injury, and can also be used to monitor

response to rejection therapy. cfDNA sequencing can

also be used to monitor changes in the microbiome after

transplantation, in response to alterations in immunosuppression and antimicrobial therapy, and can potentially be

used to diagnose opportunistic infections. Finally, quantification of

anellovirus cfDNA load offers the possibility of assessing the overall

status of the host immune response, which is a necessary first step for

providing individualized immunosuppressive therapy to transplant

recipients.

CELL-FREE DNA IN PRENATAL MEDICINE

Fetal genetic disorders are among the leading causes of stillbirth, neonatal death, and long-term developmental delay. Prenatal diagnosis

of fetal genetic conditions relies on invasive sampling of fetal tissues

via either amniocentesis (sampling of amniotic fluid) or chorionic

villus sampling (CVS; sampling of placental tissue). Prenatal diagnostic

3844 PART 20 Frontiers

tests are invasive, ultrasound-guided, needle-based procedures that

unfortunately carry a small risk of miscarriage, preterm rupture of the

membranes, and preterm birth. Historically, safer noninvasive screening for fetal genetic disorders, particularly aneuploidy (chromosomal

copy number abnormalities such as trisomy 21 or Down’s syndrome),

was performed by combining maternal serum analyte levels (e.g.,

pregnancy-associated plasma protein A [PAPP-A], β-human chorionic

gonadotropin [β-hCG], α-fetoprotein [AFP], inhibin) with sonographic assessment of the fetal nuchal translucency (a measurement of

the distance between the scalp and skin in the area of the fetal neck).

While traditional prenatal screening is safer than invasive diagnostic

tests, these modalities suffer from lower detection rates and higher

false-positive rates. Recent advances in cell-free fetal DNA technology

have introduced safe noninvasive prenatal tests with much higher

accuracy than traditional screening. This section describes the use of

cell-free fetal DNA testing in prenatal medicine.

■ NONINVASIVE PRENATAL TESTING

FOR FETAL ANEUPLOIDY

Passage of fetal cells into the maternal bloodstream was initially

described almost 80 years ago. A mismatch between fetal red blood

cells and maternal red blood cells was subsequently identified as the

cause of isoimmunization and hydrops fetalis, a potentially lethal

condition in which maternal antibodies develop against “foreign” fetal

red blood cell antigens. With the development of Rho(D) immune

globulin (RhoGAM) and its use to prevent RhD isoimmunization in

RhD-negative mothers carrying RhD-positive pregnancies, attention

then turned toward targeting fetal red blood cells for the detection of

fetal aneuploidy. This is because fetal red blood cells contain a nucleus

and genomic material, whereas maternal red blood cells do not. Unfortunately, initial research endeavors in the area failed to yield significant

results due to several inherent limitations. First, the number of fetal red

blood cells circulating among the more prevalent maternal red blood

cell background is quite small. Second, the half-life of the circulating

fetal red blood cells in maternal serum was found to be quite long,

and circulating cells may actually originate not just from the current

pregnancy but even from a prior pregnancy. Finally, the technology

available at the time, namely fluorescent-activated cell sorting, was

unable to accurately separate the cells to isolate the fetal cells from the

maternal background.

In 1990, Lo and colleagues determined that cell-free fetal DNA, and

not just fetal cells, crosses the placenta and could serve as a target for

fetal aneuploidy detection. This was subsequently confirmed when the

presence of DNA from the Y chromosome was found in the serum of

pregnant women carrying a male fetus. Over the next two decades,

cfDNA circulating in maternal blood was characterized, and different

genetic techniques were employed to delineate differences between

cell-free fetal DNA and cell-free maternal DNA. Cell-free fetal DNA

has several advantages as a diagnostic target over circulating fetal cells.

First, the fraction of fetal DNA is much higher (~5–20%) than the fraction of fetal cells in maternal circulation. Second, cell-free fetal DNA

declines within days of delivery, thereby reducing the risk of DNA from

previous pregnancies leading to erroneous results.

Unfortunately, in the case of fetal aneuploidy detection, the actual

sequence of fetal DNA from the chromosome of interest is identical

to the maternal DNA sequence, and the challenge is to distinguish one

from the other. In the early 2000s, the most promising technique for

distinguishing fetal from maternal DNA was leveraging epigenetic differences between the two, particularly in the chromosomes of interest,

namely chromosome 21 and chromosome 18. In 2008, two landmark

studies, one by Chiu and colleagues and one by Fan and colleagues,

described the utility of massive parallel sequencing technology as a

counting method for noninvasive detection of fetal aneuploidy. This

sequencing technology allowed for a comparison of the number of

sequencing reads from chromosome 21 against an expected reference

number of reads. In other words, if more sequences from chromosome

21 were found in maternal serum than what one would expect, the

fetus was suspected of having Down’s syndrome. While these findings

were based on analysis of limited cohorts, subsequent trials in more

robust and diverse cohorts confirmed these findings, and cfDNA testing became an approved clinical test in 2011.

While the initial focus of cfDNA testing was on pregnant women at

high risk for fetal aneuploidy (e.g., advanced maternal age, high-risk

status based on serum analyte screening, presence of fetal structural

anomalies), subsequent studies confirmed the utility of cfDNA technology in both high-risk and low-risk cohorts, opening the path for

widespread adoption. Today, the detection rate for the common aneuploidies (trisomies 21, 18, and 13) and for sex chromosome aneuploidy

using cfDNA testing is >98% (trisomy 21 detection as high as 99.8%),

and cfDNA testing is offered by both private companies and several

public nationwide screening programs. The positive predictive value of

these tests is variable and dependent on a priori maternal risk; therefore, most societal guidelines recommend confirmatory/diagnostic

testing using amniocentesis or CVS in cases of positive cfDNA testing.

■ NONINVASIVE PRENATAL SCREENING VERSUS

NONINVASIVE PRENATAL DIAGNOSIS VERSUS

NONINVASIVE PRENATAL TESTING

When data from clinical trials proved the improved performance of

cfDNA testing over traditional screening, it became unclear whether to

classify the test as noninvasive prenatal screening (NIPS) or noninvasive

prenatal diagnosis (NIPD). On the one hand, cfDNA testing leverages

circulating segments of fetal DNA in maternal serum, unlike traditional

screening, which is dependent on indirect analytes such as PAPP-A

or β-hCG. On the other hand, variable positive predictive values for

different aneuploidies still did not reach the diagnostic accuracy of

amniocentesis or CVS, and most cfDNA segments originate from the

placenta, rather than from the fetus. Therefore, over time, the label of

noninvasive prenatal testing (NIPT) became commonly used in both

the scientific literature and the lay literature. Today, much confusion

remains for both patients and providers as to the accuracy of these tests

and their inherent limitations.

■ DEPENDENCE ON FETAL FRACTION

As clinical experience with NIPT grew, it became clear that the accuracy of the test depends on the fraction of fetal DNA in maternal

serum. While fetal DNA can be found in maternal serum as early as

4–5 weeks’ gestation, the fetal fraction increases with gestational age,

and most NIPT programs recommend testing after 10 weeks’ gestation.

Although each NIPT company or national NIPT program uses a different fetal fraction cutoff to accurately report a result, a small percentage

of patients will receive a “no call” or uninterpretable result due to low

fetal fraction. Factors leading to “no call” also include early gestational

age (usually <10 weeks), maternal obesity, underlying maternal autoimmune disease or malignancy, and vanishing twin (twins in which

one fetus demises early in gestation). Several studies suggested an association between a no-call result and increased risk for aneuploidy, and

the appropriate management of a no-call result remains controversial,

with some recommending a repeat NIPT test, while others recommend

invasive diagnostic testing given the increased aneuploidy risk.

■ NIPT IN MULTIPLE GESTATIONS

While initial results questioned the utility of NIPT in multiple gestations, more recent reports from large cohorts suggest that the aneuploidy detection rate in twins mirrors that in singletons. Moreover,

many NIPT programs report zygosity in cases of multiple gestations.

This is particularly important since one would assume that NIPT

results from monozygotic twins (twins originating from a single

embryo) would indicate the presence of euploidy or aneuploidy in

both twins, and zygosity testing is also important when the number of

placentas has not been determined or remains unclear based on sonography. Monochorionic twins (all of whom are monozygotic twins)

are considered higher risk than dichorionic twins (most of whom are

dizygotic twins but can also be monozygotic twins) and are managed

differently. Newer NIPT platforms can also distinguish the presence of

aneuploidy in a single fetus in cases of dizygotic twins. NIPT is still not

validated for triplet gestations or multiple gestations in which a single

fetal reduction or fetal loss occurs.

3845Circulating Nucleic Acids as Liquid Biopsies and Noninvasive Disease Biomarkers CHAPTER 490

■ NIPT FOR FETAL MICRODELETION/

DUPLICATION SYNDROMES

Advanced molecular technologies, particularly chromosomal microarray (CMA), are now the preferred diagnostic method when performing amniocentesis or CVS, particularly in the setting of a known fetal

anomaly or fetal growth restriction. Unlike traditional karyotype,

which relies on the visual assessment of chromosomes by a cytopathologist with a resolution of ~5–10 MB, CMA is an automated molecular

technique that can identify differences as small as a single nucleotide.

The main benefit of CMA over traditional karyotype is in its ability to

detect copy number variants such as microdeletion/duplication syndromes (e.g., 22q11 microdeletion or DiGeorge’s syndrome). Although

the focus of NIPT was initially on the prenatal detection of Down’s syndrome and other aneuploidies, the potential for NIPT to detect smaller

chromosomal rearrangements, including copy number variations

(CNVs), was recently evaluated and introduced into clinical care. This

is important because CNVs carry a significant neonatal disease burden, and the risk of such variants may be more common than Down’s

syndrome among young pregnant women. Unfortunately, the detection rates for such conditions by NIPT are not as robust as they are

for aneuploidy detection, and the positive predictive values are lower.

Moreover, because most cell-free fetal DNA originates from the placenta, there is concern that confined placental mosaicism may result in

erroneous NIPT results that target smaller genomic aberrations. High

false-positive rates may lead to unnecessary parental anxiety, increased

rates of invasive confirmatory testing, and pregnancy termination.

■ WHOLE GENOME NIPT AND

SINGLE-GENE DISORDERS

Most NIPT platforms use next-generation sequencing to obtain DNA

reads covering the entire genome but only report a fraction of the

sequences analyzed. This has led to a debate as to whether whole

genome NIPT should be offered and reported. As with any robust and

comprehensive test, questions regarding false-positive rates and abnormalities of unclear significance are being considered. Also, whether to

limit the results to known disease-causing mutations remains controversial. While whole genome NIPT is being considered, NIPT platforms targeting single-gene disorders have been recently developed,

and these are slowly becoming clinically available. This is a significant

advance because maternal carrier screening of single-gene disorders

is becoming more prevalent, with clinically available comprehensive

gene panels covering hundreds of potential disease loci. In the absence

of NIPT, couples deemed at risk based on maternal and/or paternal

carrier screens (e.g., an asymptomatic couple in which both parents are

heterozygous for an autosomal recessive gene mutation such as cystic

fibrosis) would need to undergo an invasive diagnostic procedure in

order to confirm or exclude a significant fetal condition (homozygous for

an abnormal cystic fibrosis gene).

■ NIPT USING CELL-FREE RNA

Cell-free fetal DNA is not the only nucleic acid found in maternal

serum. Cell-free fetal RNA is also present, providing a unique snapshot

of which genes are being transcribed at any given time in pregnancy.

Cell-free fetal RNA can also be linked to tissue-specific genes, offering

a unique window into developing fetal organs. Moreover, serial analysis

of the pregnancy transcriptome and comprehensive analysis of cfRNA

in maternal serum over time have been studied as means of early

detection of common third-trimester obstetric complications including

preeclampsia, fetal growth restriction, and preterm birth.

■ NIPT DETECTING MATERNAL MALIGNANCY

Maternal malignancy complicates approximately 1 in 1000 to 1 in

1500 pregnancies. Common cancers include hematologic, breast,

and gynecologic cancers, including cancers of the cervix and ovary.

Because next-generation technologies sequence both maternal and

fetal DNA, changes in DNA levels may not only indicate fetal disorders

but also maternal disease. While most uninterpretable NIPT tests result

from low fetal fraction or other fetal and/or placental conditions, a

small subset may be due to undiagnosed maternal malignancy. Cases

of maternal lymphoma, leukemia, gastrointestinal malignancy, and

neuroendocrine tumors detected incidentally by NIPT have been

described in the literature.

■ BEYOND NEXT-GENERATION SEQUENCING

While most NIPT platforms use next-generation sequencing technology, some studies have analyzed the utility of digital PCR and fetal

cell analysis for NIPT. Digital PCR has been described as a useful tool

for testing pregnancies at risk of X-linked and autosomal recessive

single-gene disorders for both single mutations and compound heterozygous mutations. Although original attempts to isolate circulating fetal

cells were abandoned >20 years ago, recent advances have led to new

technologies targeting both circulating fetal nucleated red blood cells

and trophoblasts. This progress, still in the research phase, suggests that it

may be possible to identify and isolate single cells as early as the first trimester and analyze them by whole-gene genome amplification followed

by copy number analysis using arrays or next-generation sequencing.

SUMMARY

Collectively, the uses described in this chapter summarize some of

the myriad clinical applications of liquid biopsy methods including

cfDNA profiling in oncology, transplantation medicine, and obstetrics.

For each area, several existing assays have enabled transformative

approaches to noninvasive diagnosis of disease, assessment of associated risk, and assessment of therapeutic responses. For example,

several tumor-derived cfDNA assays are currently available for early

cancer detection, noninvasive genotyping, response assessment, and

MRD monitoring across diverse cancer types. Liquid biopsy assays for

oncology can provide highly sensitive measures of tumor burden and

associated risk and can also be used to monitor response to anticancer

therapies. Prominent among applications of liquid biopsies is the detection of MRD following definitive therapy for solid cancers as a strongly

prognostic biomarker with very high positive predictive value for

risk of recurrence. Similarly, cell-free nucleic acids have transformed

transplantation medicine. Rising levels of donor-derived cfDNA can

indicate acute allograft rejection, and clinical dd-cfDNA assays are

now being used for noninvasive rejection surveillance in lieu of invasive biopsy procedures. Microbial cfDNA analysis can potentially be

used to screen for opportunistic infections after transplantation and

for the hypothesis-free diagnosis of posttransplant infections, and

such clinical assays are currently in development. Moreover, quantification of particular microbial cfDNA sequences, such as that of the

Anelloviridae family, could one day enable transplant physicians to

assess the overall status of the alloimmune response—a goal that could

enable personalization of immunosuppressive therapy to maximize

efficacy and to prevent the long-term toxicities associated with these

medications. Finally, detection of cell-free fetal or placental DNA has

improved the prenatal detection of fetal aneuploidy and other genomic

abnormalities. Accordingly, the potential of these liquid biopsy techniques is rapidly expanding and places us at the cusp of a revolution in

personalized management of patients across medical disciplines using

innovative noninvasive techniques.

■ FURTHER READING

Alix-Panabières C, Pantel K: Clinical applications of circulating

tumor cells and circulating tumor DNA as liquid biopsy. Cancer

Discov 6:479, 2016.

Bianchi DW, Chiu RWK: Sequencing of circulating cell-free DNA

during pregnancy. N Engl J Med 379:464, 2018.

Chabon JJ et al: Integrating genomic features for non-invasive early

lung cancer detection. Nature 580:245, 2020.

Crowley E et al: Liquid biopsy: Monitoring cancer-genetics in the

blood. Nat Rev Clin Oncol 10:472, 2013.

De Vlaminck I et al: Temporal response of the human virome to

immunosuppression and antiviral therapy. Cell 155:1178, 2013.

Khush KK: Clinical utility of donor-derived cell-free DNA testing in

cardiac transplantation. J Heart Lung Transplant 40:397, 2021.

Kurtz DM et al: Circulating tumor DNA measurements as early

outcome predictors in diffuse large B-cell lymphoma. J Clin Oncol

36:2845, 2018.

3846 PART 20 Frontiers

Kurtz DM et al: Dynamic risk profiling using serial tumor biomarkers

for personalized outcome prediction. Cell 178:699, 2019.

Moding EJ et al: Detecting liquid remnants of solid tumors: circulating

tumor DNA minimal residual disease. Cancer Discov 2021.

Morain S et al: A new era in noninvasive prenatal testing. N Engl J

Med 369:499, 2013.

Warsof SL et al: Overview of the impact of noninvasive prenatal testing on diagnostic procedures. Prenat Diagn 35:972, 2015.

Many hundreds of human diseases, collectively known as protein conformational diseases or protein folding disorders, result from protein

misfolding due to intrinsic and extrinsic errors amplified by exposures

to environmental and physiologic stress conditions. Despite many

years of considerable effort, there remains no useful algorithm that can

effectively predict protein tertiary (three-dimensional) structure from

primary amino acid sequence (and its variants), let alone posttranslationally modified (from environmental exposures) primary sequence.

Such events challenge the integrity of the proteome and can lead to

premature clearance, mislocalization, dysfunction or aggregation of

proteins, thus affecting cellular robustness, health, and longevity. Mismanagement of the proteome is the basis of a broad class of hundreds

of diseases that include orphan lysosomal storage diseases, type 2

diabetes, cystic fibrosis, certain fibrotic diseases, metabolic diseases,

muscle wasting diseases, cancer, and neurodegenerative diseases, as

exemplified by Alzheimer’s disease, frontotemporal dementia, Parkinson’s disease, amyotrophic lateral sclerosis (ALS), and Huntington’s

491 Protein Folding Disorders

Richard I. Morimoto, G. Scott Budinger

Eye Neuronal tissue

Lung

Pulmonary alveolar

proteinosis

Cystic fibrosis

Thyroid

Medullary thyroid carcinoma

Immune system

Systemic AL amyloidosis

Multiple myeloma

Cataracts

Corneal lactoferrin

amyloidosis

Hereditary lattice

corneal dystrophy

Alzheimer’s disease

Amyotrophic lateral sclerosis

Familial British dementia

Familial Danish dementia

Parkinson’s diease

Huntington’s disease

Muscle

Inclusion body

myositis/myopathy

Aortic medial

amyloidosis

Cardiac amyloidosis

(e.g., transthyretin

cardiomyopathy)

Liver

Systemic diseases

Type 2 diabetes mellitus

α1 Antitrypsin deficiency

Lysozyme storage diseases

p53-dependent cancers

Pancreas/islet a cells

FIGURE 491-1 Diseases of protein folding. A representative list of tissues and known folding diseases.

disease (Fig. 491-1). For each of these diseases and many others

described in this textbook, aging is the major contributing risk factor.

The challenge at the biochemical and molecular level is for the cell

to achieve a stable and functional proteome during development that

persists through young adulthood and to maintain it throughout aging.

For humans, this is necessary for the operational health of each of the

tens of trillions of cells that compose our ~80 organs for health span

and life span. To achieve this goal, our cells have evolved a remarkably

efficient proteostasis network (PN) composed of molecular chaperones

and other highly conserved components essential for normal protein

synthesis, folding, translocation, and degradation (Fig. 491-2) that

balances input with output and that ensures every protein is functional.

The PN is essential for all tissues and for the diverse protein-protein

interactions required for cell signaling, biosynthetic processes, and

the structural demands for tissue shape, mechanical properties, and

function. An equal, if not more important, role for the PN is to detect,

prevent, and remove misfolded and aggregated proteins that accumulate in stress, aging, and disease and that thereby interfere with cellular

health. Understanding how proteostasis is achieved and maintained is

of fundamental biological interest and essential to prevent age-associated

protein folding disorders.

All organisms use an evolutionarily conserved set of molecular

machines for the synthesis, folding, transport, and removal of unnecessary and damaged proteins. The PN evolutionarily has adapted to the

highly specific physiologic requirements of tissues and the expression

of abundant and rare proteins with wide-ranging solubilities, folding

requirements, stabilities, and structural demands. Added to this complexity of natural clients for the PN is the additional load generated by

genetic mutations carried in natural populations and environmental

stressors that challenge the capacity of the PN. Despite the central role

of proteins as the workhorse of the cell, they are highly susceptible to

molecular damage, whether due to intrinsic metastability or genetically

inherited mutation or because of error-prone synthesis. Hence, dysfunction in the PN may clinically manifest as either a gradual decline

in homeostatic function as occurs with genetic mutations or a loss of

resilience in the face of environmental stressors. Thus, clinicians see the

consequences of proteostasis failure and cellular dysfunction in both the

myriad disorders that present to them as

age-associated clinical problems and the

increased morbidity and mortality associated with trauma, infection, and other

acute illness requiring hospitalization in

older individuals.

PROTEIN QUALITY

CONTROL MECHANISMS

The PN monitors and controls the flux

of protein synthesis in order to promote functional folding and to minimize the accumulation of off-pathway

aggregation-prone intermediates by

their selective disaggregation or degradation. However, unlike an automobile

assembly plant for which each item is

designed and engineered for assembly

into functional components, the properties of the PN tolerate the tremendous

chemical noise and diversity generated

by coding region polymorphisms and

biosynthetic errors among its client

components, while maintaining the

ability to recognize and remove kinetically unstable conformational states of

proteins that would compromise assembly and function. Proteins are highly

sensitive to fluctuations in their intracellular environment caused by shifts in

energetics, pH, oxidizing and reducing

conditions, and the myriad of small

3847Protein Folding Disorders CHAPTER 491

Consistent with the recognition that the formation of off-pathway

aggregates is a kinetic component of proteostasis are the concerted

activities of disaggregase machines that can unravel protein aggregates

into the unfolded, extended polypeptide chain. These disaggregases

correspond to the AAA+ proteins, corresponding to ClpB in bacteria

and Hsp104 in yeast and plants and functionally related to a redirected

HSP70 machine in other eukaryotes and metazoans that interacts with

HSP110 and specific combinations of J-domain proteins capable of

inducing disaggregation.

The subcellular organelles, the mitochondria, and endoplasmic

reticulum (ER) are responsible for ~20% of the proteome. Chaperone

interactions are essential to maintain the extended polypeptide chain

in a recognition-competent state for the respective organellar receptors required for translocation across membranes. Upon import, each

translocated polypeptide is met by organellar-specific chaperones of the

HSP70 and J-domain family for folding and assembly. While the mitochondrial genome encodes 13 proteins required for electron transport,

the great majority of mitochondrial proteins are encoded by the nuclear

genome, synthesized in the cytosol, and then imported across the outer

and inner mitochondrial membranes. Hence, maintenance of the mitochondrial proteome relies on both the cytosolic and mitochondrial PN.

For translocation into the lumen of the ER, the extended polypeptide

interacts with a set of glycosyltransferases, calnexins, calreticulins and

disulphide isomerases. Proteins that misfold in the ER are recognized

and retrotranslocated to the cytoplasm, where they are directed to the

ubiquitin-proteasome system (UPS) for unfolding and degradation.

The PN is balanced by the essential catabolic processes of the

ubiquitin-proteasome and the autophagy-lysosomal machines that

recognize proteins for degradation and recycling. The UPS is generally

considered the primary pathway by which most proteins are recognized

and tagged for degradation; the autophagy-lysosome system is highly

responsive to nutrient changes and damage, recognizing and engulfing

organelles and other subcellular compartments and large aggregates

and inclusions. In addition to their role in the regulated turnover of

cellular proteins, these degradation systems are essential for protein

quality control and for limiting the accumulation of misfolded and

aggregated proteins during stress conditions, aging, and disease.

Protein turnover by the UPS involves an enzymatic cascade of E1,

E2, and E3 enzymes that utilize ubiquitin to tag clients, followed by

degradation of the polyubiquitinated substrates by the 26S proteasome.

Client specificity involves the large family of ~600 ubiquitin ligases.

In addition to their roles to mark proteins for degradation, the ubiquitination machinery has numerous additional functions in cellular

processes. For example, the ubiquitin ligase listerin is associated with

the ribosome and ubiquitinates nascent chains that stall in translation

to prevent the accumulation of aberrant polypeptides that would subsequently aggregate. Ubiquitination of these nonnative clients by the

ubiquitin ligase activity of the co-chaperone CHIP is central to the triage decision of the HSP70/HSP90 complex between client folding and

proteasome-mediated degradation. ER-targeted clients that are misfolded are retrotranslocated to the cytoplasm, polyubiquitinated, and

degraded by cytosolic proteasomes in a process termed ER-associated

degradation (ERAD). Ubiquitination also provides cross-talk between

the proteasome and autophagy pathways by targeting clients for lysosomal degradation and for endosomal sorting. Chaperones co-label

a protein as nonnative, recruiting other proteins that place ubiquitin

chains on the damaged protein for degradation by the proteasome.

Alternatively, chaperones can label proteins or protein aggregates to

target them to the lysosome. In this process, damaged proteins are

degraded by the lysosome, an intracellular organelle with an acidic

environment enriched in proteases, through autophagy.

While there is a good general understanding of the process of in vivo

chaperone-dependent protein folding, the details of how decisions are

made for each client in the cell—whether and for how long to be maintained in a nonnative folding state through chaperone interactions in

a nucleotide-independent state, or to assemble into a stable chaperone

complex for subsequent assembly into a functional state, or to interact

with chaperones to fold directly to a native state—remain to be fully

addressed.

molecules and metabolites that affect folding and function. Added to

this range of perturbations are the effects of external stress caused by

elevated temperatures, infections, abnormal redox environments, or

osmolytes that can have profound consequences on protein-folding

thermodynamics, kinetics, and function. These intracellular and extracellular stress conditions, if not properly addressed, would be predicted

to amplify further protein instability from sequence polymorphisms

and biosynthetic errors that contribute to the stress of protein misfolding. The PN is organized at the cellular level into a series of highly

coordinated molecular machines that direct the expression, biogenesis,

and functional health of essentially all proteins (Fig. 491-2). Serving

more than as regulator and orchestrator of these highly synchronized

events, the PN is essential for protein quality control and prevention

of the appearance of off-pathway conformational states and accumulation of aggregates and amyloid species. Proteome health involves the

constant exchange between the intrinsic physicochemical properties of

polypeptides and the biological milieu of the cellular environment in

which protein sequences and function evolved.

Beginning with the synthesis of the nascent polypeptide on the

ribosome, ribosome quality control (RQC) factors and cytoplasmic

chaperones of the HSP70, HSP90, DNAJ/HSP40, chaperonin/HSP60,

and small heat shock proteins (sHSPs) family ensure co-translational

and posttranslational folding for the cell. Approximately 60% of the

proteome resides in the cytoplasm and nucleus, for which the RQC,

HSP70, HSP90, and HSP60 chaperones regulate the folded state of

client proteins, together with co-chaperones, through cycles of ATP

binding and hydrolysis. Chaperones of the HSP70 and J-domain family

are particularly well studied for their ability to interact transiently with

nascent polypeptides in protein synthesis through short, dispersed,

hydrophobic regions using the energy from nucleotide hydrolysis to

stimulate the release of partially folded intermediates that either reenter the chaperone cycle or are in a folded native state. For other clients,

such as transcription factors, kinases, phosphatases, and signaling

molecules, folding to the functional state is regulated and dependent

upon interactions with the HSP90 chaperone and other regulatory

co-chaperones to form stable heteromeric complexes to maintain

the client in a partially folded state primed for subsequent regulated

release.

Intermediate Native state

folded states

Misfolded states

Molecular

chaperones

Chaperones

Autophagy

Proteasome

Unfolded nascent

polypeptide

Normal turnover

Improper trafficking Toxic folds Degradation

Emphysema Amyloidoses Cystic Fibrosis

Cystic fibrosis

transmembrane

conductance regulator

Abeta, tau, Huntington,

SOD1, α-synuclein

α1 Antitrypsin

FIGURE 491-2 The proteostasis network and protein folding diseases. The process

of protein biogenesis involves the action of molecular chaperones to ensure the

transition of the unfolded nascent polypeptide through intermediates to the folded

native state. Such proteins then have a natural turnover. Off-pathway species are

prevented by the actions of chaperones and the recognition of nonnative misfolded

states and aggregates by the autophagic-lysosomal machinery and the ubiquitinproteasome. When misfolded species escape quality control, they can then become

improperly trafficked, as occurs for α1

 antitrypsin associated with emphysema; for

toxic folds as occurs for amyloid beta, tau, huntington, and SOD1 in amyloidogenic

neurodegenerative diseases; or prematurely degraded as occurs for cystic fibrosis

transmembrane conductance regulator (CFTR) associated with cystic fibrosis.

3848 PART 20 Frontiers

CELL STRESS RESPONSES: SENSORS AND

REGULATORS OF PROTEIN DAMAGE

Cell stress responses are ancient genetic networks that detect, adapt to,

and protect all cells against toxic environmental stressors and physiologically relevant changes in their cellular environment (Fig. 491-3).

At the core of these cell stress responses are molecular switches: (1) the

heat shock response (HSR) that protects proteins in the cytoplasm and

nucleus controlled by HSF-1; (2) the unfolded protein response (UPR)

of the ER (UPRER) controlled by XBP1, ATF6, and ATF4; (3) the UPR

of the mitochondria (UPRmito) controlled by ATFS1; (4) the DAF-16/

FOX-O stress response pathway associated with insulin signaling; and

(5) the antioxidant stress response regulated by NRF-2. Collectively,

these cell stress responses and their respective transcription factors are

each essential for all cells and tissues and are regulated both autonomously and cell nonautonomously across tissues in metazoans to detect

proteotoxic stress and to adapt to and protect the cell against the toxic

consequences of protein damage. While each of these cell stress pathways can be activated independently, they are also induced in different

combinations according to the chemical and physiologic properties of

the stress signal(s) and provide cross-protective mechanisms.

The HSR is an evolutionarily conserved cellular defense mechanism

that protects cells against proteotoxicity associated with misfolding,

aggregation, and proteome mismanagement. HSF-1 inducibly regulates

transcription of genes encoding molecular chaperones and components of the PN. In unstressed cells, HSF-1 is cytoplasmic or nuclear

and exists in an inert monomeric state negatively regulated by weak

interactions with the molecular chaperones HSP70 and HJD-1. Upon

heat shock stress, HSF-1 rapidly trimerizes to acquire DNA-binding

activity and undergoes extensive posttranslational modifications, binds

to heat shock elements in promoters of stress responsive genes, and

relocalizes into nuclear stress bodies. Upon dissipation of the stress

signal, attenuation of the HSR involves the active repression of HSF-1

DNA binding through acetylation and loss of HSF-1 transcriptional

activity by reassociation with HSP70 and other molecular chaperones

and with HSBP1, leading to dissociation to the monomeric inert state.

In addition to HSF-1 being essential for the HSR and cell and organismal stress resilience, HSF-1 is essential during early development in

metazoans, functions as a maternal factor for gametogenesis, regulates

oocyte maturation by activating genes that function in the meiotic cell

cycle, is constitutively activated in cancer, and is necessary to maintain

NAD+ and ATP levels.

In the ER, the UPRER involves three stress arms regulated by the TFs,

XBP1, ATF6, and ATF4, that bind to specific cis-elements for these

ER-stress-responsive arms. XBP1 is activated by the transmembrane

endoribonuclease IRE1, which is a transmembrane protein with kinase

and endoribonuclease (RNase) activity that senses misfolding in the

ER directly, leading to autophosphorylation, oligomerization, and

acquisition of RNase activity. These events allow active IRE1 to cleave

XBP1 mRNA, generating a spliced transcript (XBP1s) that encodes

XBP1, which induces the transcription of UPR target genes. ER stress

also promotes the relocalization of ATF6 from the ER membrane to the

Golgi apparatus, where it is cleaved by the proteases SP1 and SP2, generating a cytosolic fragment of ATF6 that translocates to the nucleus

to direct transcription of a complementary set of UPR genes. Together,

XBP1 and ATF6 induce the expression of genes involved in protein

folding, ER-associated protein degradation, and lipid metabolism

pathways. A third ER transmembrane protein, PERK, also induced by

ER stress, promotes translation of the TF ATF4 by phosphorylating

the translation initiation factor eIF2α. Under these conditions, ATF4

mRNA is preferentially translated, leading to selective expression of

the proapoptotic TF CHOP, which elicits apoptosis in cells in which

ER stress is not resolved, presumably to remove damaged cells from

the population.

For mitochondria, the UPRmito response involves ATFS1, which

contains a mitochondrial targeting sequence and a nuclear localization signal. Under normal cellular conditions, ATFS1 is imported into

mitochondria and degraded, but upon mitochondrial stress, ATFS1 is

directed only to the nucleus to regulate transcription of genes encoding mitochondrial chaperones, mitochondrial import machinery, and

glycolysis. In mammals, the UPRmito is regulated by ATF5, which corresponds to ATFS-1 in Caenorhabditis elegans.

In metazoans, the integration of stress survival strategies includes

the antioxidant factor SKN-1/NRF2, the insulin-signaling factor DAF16/FOXO, and the tissue identity factor PHA-4/FOXA. Oxidative

and xenobiotic stresses activate OxR, which controls the expression

of redox-regulatory proteins and components of protein degradative pathways mediated in mammals by NRF1/NFE2L1 and NRF2/

NFE2L2, which corresponds to SKN-1 in C. elegans. NRF1 is an

ER-resident factor that undergoes regulated proteolytic cleavage upon

activation to control expression of genes encoding subunits of the proteasome and the UPS. NRF2 in the cytoplasm is negatively regulated by

the redox-sensitive ubiquitin ligase KEAP-1; consequently, inactivation

of KEAP-1 by oxidative and electrophilic stress leads to stabilization

and nuclear translocation of NRF2, which, in turn, induces the expression of antioxidant proteins and detoxification enzymes.

ORGANISMAL PROTEOSTASIS IN

AGING AND DISEASE

Much of our understanding on protein quality control mechanisms

has come from in vitro studies with purified molecular chaperones

or components of the UPS, complemented with cell extracts and cellbased assays using yeast or mammalian cells grown in tissue culture.

A common theme that emerges from these studies is that of hormesis,

in which transient bouts of activation of the HSR, UPRER, or UPR mito

improves lifespan and organismal resilience but chronic activation is

detrimental.

The importance of these pathways is further highlighted by studies

suggesting neuronal coordination of stress responses at the level of the

organism. When neuronal mechanisms fail, the HSR reverts to cell-autonomous control. At the organismal level, however, the HSR, UPRER,

and the UPR mito in C. elegans are regulated by cell-nonautonomous

control through neuronal signaling. When neuronal signaling is

impaired, the HSR reverts to cell-autonomous control. Neuronal signaling also regulates the UPR mito with disruption of mitochondrial

function in C. elegans neurons activating the UPR mito in nonneuronal

tissues, supporting a role for a mitokine signal. Perturbation of the

mitochondrial electron transfer chain (ETC) was shown to increase

life span in both invertebrates and rodents through the activation of

the UPR mito. The response to mitochondrial dysfunction in C. elegans depends upon the severity of mitochondrial impairment with a

mild reduction of ETC having hormetic effects on organismal stress

resilience, proteostasis, and longevity by resetting the cytoplasmic

High

Risk

Development Reproduction Aging Disease

Programmed

Repression of

the Heat Shock

Response, UPR,

and Oxidative

Stress Response

Stress responses

Molecular chaperones

Protein quality control

Proteostasis

Low

Risk

FIGURE 491-3 Aging and protein folding diseases. Aging is the major risk factor

for degenerative diseases. Cell stress responses (heat shock response and the

unfolded protein responses in the endoplasmic reticulum and mitochondria) decline

at reproductive maturity in studies from Caenorhabditis elegans and prevent

adaptive and protective increased expression of molecular chaperones to prevent

protein misfolding. UPR, unfolded protein response.

3849Protein Folding Disorders CHAPTER 491

HSR through HSF-1, independent of ATFS-1 and the UPR mito.

Mild perturbation of the ETC in Drosophila muscle also has systemic

benefits on organismal health and life span involving the insulin signaling. Communication between neurons also regulates the UPRER

in peripheral tissues of C. elegans. During infection of C. elegans by

pathogens, induction of the UPRER in nonneuronal tissues is mediated

by sensory neurons, suggesting an organismal stress response. Cellnonautonomous regulation of the UPRER has also been observed in

mice, where overexpression of active XBP1 in pro-opiomelanocortin

neurons activates the UPRER in the liver.

Other forms of intertissue communication that regulate proteostasis

with beneficial effects on organismal health include: cholinergic signaling across the synaptic junction to enhance or inhibit HSF-1-regulated

proteostasis in the muscle cells of C. elegans; transcellular chaperone

signaling between somatic tissues or between somatic tissues and

neurons of C. elegans to regulate the expression of HSP90 in receiving

tissues via the tissue code factor PHA4/FOXA, protection of neurons

and glial cells from elevated temperature-induced death by overexpression of small HSPs in Drosophila flight motor muscle cells, enhancing

proteostasis in C. elegans muscle cells by elevated expression of DAF16/

FOXO in the intestine, and that overexpression of dFOXO/4E-BP

in Drosophila muscle influences proteostasis in retina, brain, and

adipose tissues to delay the age-dependent accumulation of protein

aggregates.

Cell stress responses and proteostasis decline in aging. Insights

on the relationship between proteostasis, cell stress, and aging have

come primarily from C. elegans with support from other invertebrate

and vertebrate model systems and human cells. A set of endogenous

metastable proteins that exhibit temperature-sensitive properties were

shown to misfold in C. elegans at the permissive temperature during

aging, which was associated with a decline in the HSR. The decline of

proteostasis in C. elegans aging is regulated via cell-nonautonomous

control involving germline stem cells that initiate the programmed

repression of the organismal HSR resulting in the loss of stress resilience and age-associated protein aggregation. This is regulated at

reproductive maturity by an epigenetic switch involving the timed

placement of repressive H3K27me3 chromatin marks at the heat shock

genes, causing chromatin inaccessibility for HSF-1. This age-dependent

decline in the HSR can be reversed either by blocking the signal from

germline stem cell signal(s) or preventing the epigenetic repressive

marks. The relationship between reproduction and inducibility of

the HSR observed in animals at reproductive maturity suggests that

the age-associated events of cellular failure and loss of tissue robustness

during aging are not random processes but, rather, highly regulated,

perhaps to ensure that somatic tissues are programmed to decline after

reproduction, consistent with the germline soma theory of aging.

Proteostasis is one of the pillars of gerontologic biology, which,

together with genomic instability, telomere attrition, epigenetic alterations, deregulated nutrient sensing, mitochondrial dysfunction, cellular

senescence, stem cell exhaustion, and altered intercellular communication, provides a mechanistic basis for the biology of aging. The programmed decline of proteostasis in early adulthood would suggest that

failure in protein quality control would have negative consequences on

the other key elements of gerontologic biology. Whether proteostasis

collapse is the first to fail or among the earliest events to fail in aging,

it is consistent with the very large number of human degenerative diseases in aging associated with protein misfolding.

PROPERTIES OF PROTEIN

FOLDING DISEASES

The complexity that arises with protein folding diseases is that all

tissues are at risk and essentially all proteins are at risk for misfolding.

Added to this is the effect of aging and that each protein folding disease

exhibits a highly variable age of onset for pathology. There is additional

complexity in classification—whether to organize folding diseases by

tissues, i.e., muscle proteinopathies or neurodegenerative diseases,

according to the specific protein that misfolds such as α1

 antitrypsin

(AAT) deficiency, or by the biophysical nature of the aggregate species

in systemic amyloidoses.

Disorders in which a specific mutation leads to protein misfolding or

the formation of a specific insoluble protein aggregates likely represent

only the tip of the iceberg of protein folding disorders. Mutations in

aggregation-prone proteins coupled with changes in the cellular environment and decline in capacity and robustness of the PN in aging

promote protein misfolding and aggregation in affected tissues. Once

aggregation has initiated, this leads to further impairment in protein

quality control pathways, causing further collateral damage and aggregation of other at-risk proteins. Such a mechanism may only manifest

clinically after a seemingly random systemic stress like pneumonia,

large bone fractures, or ischemic vascular events, possibly explaining

the rapid (1–2 years) accumulation of age-related morbidities in the

year following the first event. As such, the age-related decline in the

function of any of the components of the PN could underlie the compounding multiple morbidity that limits health span and life span in

many elderly individuals. Within this framework, it is useful to discuss

some of the better understood mechanisms of proteostasis dysfunction

that have been causally linked to diseases in humans.

■ DISORDERS THAT ENHANCE MISFOLDING AND

CAUSE PREMATURE DEGRADATION

(CYSTIC FIBROSIS)

Cystic fibrosis (CF) is a recessive disorder caused by mutations in both

alleles of the cystic fibrosis transmembrane conductance regulator

(CFTR) gene that encodes a multidomain membrane-spanning chloride ion channel protein. Thousands of mutations in CFTR have been

identified that affect CFTR biosynthesis, folding, trafficking, and function, leading to chronic obstructive lung disease, intestinal obstruction,

liver dysfunction, exocrine and endocrine pancreatic dysfunction, and

male infertility. CF is a folding disease due to its recognition by the

PN as misfolded protein. The most prominent mutation is deletion of

phenylalanine 508 (F508del), present in ~90% of CF patients. Mutant

ΔF508 retains partial channel function, but because it is recognized as

misfolded in the ER and the cytoplasm, it is marked with ubiquitin for

degradation by the ubiquitin proteasome system. Small molecules that

affect the conformation and function of mutant ΔF508 channel function can result in substantially improved outcomes in patients.

■ DISORDERS THAT INDUCE TOXIC AGGREGATES

AND LOSS OF FUNCTION (AAT DEFICIENCY)

AAT deficiency (AATD) is a co-dominant inherited disease with an

increased risk of chronic obstructive pulmonary disease (COPD), liver

disease, and vascular inflammation. Pulmonary problems are more

frequent in adults, whereas liver and skin problems may occur in adults

and children. AAT is encoded by the SERPINA1 gene, secreted into the

circulation by the liver, and responsible for inactivating endogenous

proteases, particularly those secreted by neutrophils and other inflammatory cells in the lung. Patients with AATD harbor mutations in

SERPINA1 that cause misfolding in the ER. The two major phenotypes

resulting from this abnormality highlight the diverse consequences

of misfolding on different cells and organs. In the liver, misfolding of

the mutant protein results in the formation of toxic aggregates and

hepatocyte death, manifest as liver injury and eventually cirrhosis—a

gain-of-function toxicity. In the lung, the failure to secrete sufficient

AAT causes unchecked proteolytic damage to the delicate architecture

of the alveolus, a process that is markedly worsened when neutrophils

are recruited to the lung in response to cigarette smoking. This loss-offunction phenotype manifests pathologically as emphysema and clinically as COPD.

■ INTERACTIONS WITH PN COMPONENTS THAT

CHANGE CONFORMATION, STABILITY,

OR FUNCTION (CANCER)

Mutations in the tumor suppressor p53 are among the most common

mutations observed in patients with cancer. Deletion of p53 combined

with overexpression of an oncogene is sufficient to drive metastatic